Mitogen signalling pathway protein variants

Abstract

Methods for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer are described, comprising detecting the presence of one or more specific mitogen pathway protein variants in cancer cells of the patient, wherein the presence of the one or more mitogen pathway protein variants are indicative of response to one or more of: MAP2K1 / 2 inhibitors, BRAF inhibitors, EGFR inhibitors, PI3K inhibitors, and KRAS inhibitors Related methods and products are also described.

Description

[0001]Mitogen signalling pathway protein variants Field of the Disclosure The present invention relates to methods of identifying a prognosis or treatment for a patient with cancer, based on the presence or absence of specific mitogen pathway protein variants, and particularly, although 5 not exclusively, variants in BRAF, KRAS, MAP2K1, MAP2K2, PIK3CA, AKT1 and EGFR proteins and genes encoding said mutations. Related methods, compositions and methods of treatment are also described. Background Drug resistance is a major limitation to the long-term efficacy of cancer therapies. Cancer genome 10 sequencing can delineate the genetic basis of drug resistance, but this retrospective approach requires sequencing of large numbers of post-treatment samples to distinguish causal variants from background mutational events. Despite considerable advances in the development of molecularly-targeted therapies for cancer patients, resistance to anti-cancer treatments remains a major clinical challenge (Vasan 2019). Drug resistance is 15 frequently caused by DNA single nucleotide variants (SNVs) in the cancer genome (Pao 2005), leading to point mutations in the drug target itself, or proteins within the same signalling pathway (van de Haar 2021). The study of drug resistance is crucial to understand drug mechanism of action, to generate second- generation inhibitors targeting drug-resistant proteins, for the development of combination therapies, and for effective patient stratification for second-line therapies. Current approaches often depend on sequencing 20 tumour biopsies from patients that relapse on treatment. These can be challenging samples to acquire, and so it can take years to accrue sufficient numbers to infer variant function. Moreover, these analyses are generally restricted to the most frequently observed variants, and have to be individually experimentally validated to establish a causal link to drug resistance, which is a slow process and does not allow for the direct comparison of different variant effects. Overall, these challenges limit the interpretation of cancer 25 biopsy and circulating tumour DNA sequencing data for effective patient stratification. Rapid, prospective and systematic functional annotation of variants would accelerate the discovery of drug resistance mechanisms and improve the effectiveness of cancer treatments. The present invention has been devised in light of the above considerations. Summary of the Disclosure 30 The present inventors used base editing at scale to investigate genetic mechanisms of acquired resistance to molecularly-targeted cancer therapies, identifying variants of unknown significance conferring drug resistance and drug sensitisation in cancer cells. With this approach, the inventors classify cancer variants modulating drug sensitivity into four functional classes, thus providing a systematic framework for interpreting drug resistance mechanisms. 35 The inventors have determined a landscape of genetic resistance mechanisms to 10 oncology drugs (specifically: the PARP inhibitors Niraparib and Olaparib, the allosteric MAP2K1 / 2 inhibitor Trametinib the 2 BRAF inhibitor Dabrafenib, the EGFR inhibitor Cetuximab, the EGFR TK inhibitors gefitinib and osimertinib, the pan-PI3K inhibitor pictilisib, and the KRAS G12C inhibitors sotorasib and adagrasib) from CRISPR base editing mutagenesis screens in four cancer cell lines using a guide RNA library predicted to install 32,476 variants in 11 cancer genes. By systematically analysing variant function, the inventors identify four classes 5 of protein variants modulating drug sensitivity, including variants that can be targeted with alternative inhibitors to overcome drug resistance. Single-cell transcriptomics reveals how drug resistance variants operate through distinct mechanisms, including eliciting a drug addicted cell state. Unexpectedly, we identify an EGFR variant that sensitises lung cancer cells to EGFR inhibitors. The inventors’ variant function map of genetic drug resistance mechanisms has implications for patient stratification, therapy combinations 10 and the scheduling of drugs in cancer treatment. Thus, according to a first aspect, there is provided a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer, the method comprising detecting the presence of one or more mitogen pathway protein variants in cancer cells of the patient, wherein the mitogen pathway protein variants are selected 15 from: variants that are associated with a C-terminal truncation of EGFR, and in particular a truncation after amino acids E1091 or L1038; EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, 20 E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, K499E, K499R; PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A; and AKT1 variants selected from: E17K, G16E, wherein the presence of the one or more mitogen pathway protein variants are indicative of response to one or more of: MAP2K1 / 2 inhibitors, BRAF inhibitors, EGFR inhibitors, PI3K 25 inhibitors, and KRAS inhibitors; and optionally selecting the patient for treatment or recommending treatment with one or more therapies that the patient has been identified as likely to be responsive to or one or more therapies that are not therapies that the patient has been identified as likely to be resistant to. The wording “mitogen pathway protein variants” refers to mitogen pathway proteins that have one or more of the indicated mutations. Thus, detecting the presence of one or more mitogen pathway protein in cancer 30 cells of the patient comprises detecting the presence of any one or more of the indicated mutations. Embodiments of the first aspect may have any one or more of the following optional features. In some embodiments, the variants are selected from: variants that are associated with a C-terminal truncation of EGFR after amino acids E1091 or L1038; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: I99T, L98P, S194P; MAP2K2 variants selected from: 35 Y134H; BRAF variants selected from: K499E, K499R. 3 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including KRAS E62 and / or E63 to K is predicted to be resistant to BRAF inhibitors and / or EGFR inhibitors, optionally wherein the patient is predicted to be resistant to the combination of a BRAF inhibitor and an EGFR inhibitor, optionally wherein the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is 5 cetuximab, optionally wherein the method further comprises selecting the patient for treatment with a therapy that is not a BRAF inhibitor or EGFR inhibitor, optionally a MAP2K1 / 2 inhibitor, further optionally trametinib, optionally wherein the patient who has been detected as having one or more mitogen pathway protein variants including KRAS E62 and / or E63 to K is predicted or detected as not having a mutation in the BRAF gene. 10 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including KRAS K117 to E, R and / or D119 to G is predicted to be resistant to BRAF inhibitors and / or EGFR inhibitors, optionally wherein the patient is predicted to be resistant to the combination of a BRAF inhibitor and an EGFR inhibitor, optionally wherein the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab or gefitinib, optionally wherein the method further comprises selecting the patient for 15 treatment with a therapy that is not a BRAF inhibitor or EGFR inhibitor, optionally a MAP2K1 / 2 inhibitor, further optionally trametinib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including MAP2K2 Y134 to H is predicted to be resistant to one or more of: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 inhibitors, and KRAS inhibitors, optionally wherein the BRAF inhibitor is dabrafenib 20 and / or the EGFR inhibitor is cetuximab or gefitinib and / or the MAP2K1 / 2 inhibitor is trametinib and / or the KRAS inhibitor is a KRAS G12C inhibitor or a pan-KRAS inhibitor, optionally sotorasib or adagrasib; optionally further comprising selecting the patient for treatment with one or more of: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 inhibitors, and KRAS inhibitors using an intermittent treatment scheme, or selecting the patient for treatment with a therapy that is not limited to: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 25 inhibitors, or KRAS inhibitors. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including MAP2K2 Y134 to H has also been detected as having a wild-type BRAF gene and / or expressing a wild-type B-Raf protein. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein 30 variants including MAP2K1 S194 to P is predicted to be resistant to one or more of: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 inhibitors, and KRAS inhibitors; optionally wherein the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab or gefitinib and / or the MAP2K1 / 2 inhibitor is trametinib and / or the KRAS inhibitor is a KRAS G12C inhibitor or a pan-KRAS inhibitor, optionally sotorasib or adagrasib, optionally further comprising selecting the patient for treatment with one or more of: BRAF inhibitors, EGFR 35 inhibitors, and / or MAP2K1 / 2 inhibitors, and / or KRAS inhibitors using an intermittent treatment scheme, or selecting the patient for treatment with a therapy that is not limited to: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 inhibitors, KRAS inhibitors, and / or a T-cell therapy or vaccine. In embodiments, a patient who 4 has been detected as not having mitogen pathway protein variant MAP2K1 S194 to P and / or Y134H is selected for treatment with a combination of a MAP2K1 / 2 inhibitor and a T-cell therapy or vaccine. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including MAP2K1 L98 to P and / or I99 to T is predicted to be resistant to one or more MAP2K1 / 2 5 inhibitors, optionally wherein the one or more MAP2K1 / 2 inhibitors is trametinib, optionally further comprising selecting the patient for treatment with one or more MAP2K1 / 2 inhibitors using an intermittent therapeutic scheme, and / or selecting the subject patient for treatment with a drug that is not a MAP2K1 / 2 inhibitor. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein 10 variants including a variant causing a C-terminal truncation of EGFR after amino acid E1091 or L1038 is predicted to be sensitive to one or more EGFR inhibitors, optionally wherein the one or more EGFR inhibitors are selected from: gefitinib, osimertinib, erlotinib, lapatinib, and cetuximab, optionally further comprising selecting the patient for treatment with the one or more EGFR inhibitors. In some embodiments, the variant causing a C-terminal truncation of EGFR after amino acid E1091 is a 15 single nucleotide variant (SNV), optionally wherein the SNV is a G to A mutation at location 7:55202626, or a T to C mutation at location 7:55202627, wherein the locations refer to locations in the GRCh38 reference genome assembly (NCBI RefSeq Assembly GCF_000001405.50 (GRCh38.p14)), and optionally wherein the SNV results in disruption of the splice donor site at the 5’ end of EGFR intron 27, and / or wherein the SNV results in creation of a premature stop codon. 20 In some embodiments, the variant causing a C-terminal truncation of EGFR after amino acid L1038 is a single nucleotide variant (SNV), optionally wherein the SNV is a G to A mutation at location 7:55201356, wherein the location refers to a location in the GRCh38 reference genome assembly (NCBI RefSeq Assembly GCF_000001405.50 (GRCh38.p14)), and optionally wherein the SNV results in disruption of the splice donor site at the 5’ end of EGFR intron 25, and / or wherein the SNV results in creation of a premature 25 stop codon. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including a variant causing a C-terminal truncation of EGFR after amino acid E1091 or L1038 is also detected as having an EGFR amplification mutation, an EGFR exon 19 deletion, an EGFR L858R mutation, and / or any other EGFR mutation. 30 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including a variant BRAF K499 to E or R is predicted to be resistant to one or more of: BRAF inhibitors, EGFR inhibitors, and MAP2K1 / 2 inhibitors, optionally wherein the one or more EGFR inhibitors include gefitinib, the BRAF inhibitors include cetuximab, and / or the MAP2K1 / 2 inhibitors include trametinib, optionally further comprising selecting the patient for treatment with a therapy that is not an EGFR inhibitor, 35 BRAF inhibitor or MAP2K1 / 2 inhibitor. 5 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including PIK3CA E545 to K, E542 to K, or E547 to K is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, a MAP2K1 / 2 inhibitor, an EGFR TK inhibitor, and / or a pan-PI3K inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. 5 In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, the EGFR TK inhibitor is osimertinib. In some embodiments, the pan-PI3K inhibitor is pictilisib. In some embodiments, the patient is predicted to be resistant to each of dabrafenib+cetuximab, trametinib, osimertinib and pictilisib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises 10 selecting the patient for treatment with a therapy that is not dabrafenib+cetuximab, trametinib or osimertinib, such as e.g. pictilisib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including PIK3CA E970 to G, Q969 to R and / or T972 to A is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, a MAP2K1 / 2 inhibitor, and / or a KRAS G12C 15 inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, the KRAS G12C inhibitor is sotorasib and / or adagrasib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy 20 that is not dabrafenib+cetuximab, trametinib, sotorasib, or adagrasib, such as e.g. pictilisib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including AKT1 E17 to K and / or G16 to E is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, a MAP2K1 / 2 inhibitor, and / or a pan-PI3K inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, the 25 MAP2K1 / 2 inhibitor is trametinib. In some embodiments, the pan-PI3K inhibitor is pictilisib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not dabrafenib+cetuximab, trametinib or pictilisib, such as e.g. an AKT inhibitor. In some embodiments, the AKT inhibitor is an inhibitor of AKT1, AKT2 and / or 30 AKT3, or a pan-AKT inhibitor, such as e.g. MK-2206, capivasertib, ipatasertib, afuresertib, and / or uprosertib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including MAP2K2 V195A and / or S198 to P is predicted to be resistant to a KRAS G12C inhibitor. In some embodiments, the KRAS G12C inhibitor is sotorasib and / or adagrasib. In some embodiments, the 35 patient is predicted to be resistant to each of sotorasib and adagrasib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient 6 who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not sotorasib or adagrasib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including MAP2K2 I208 to T and / or L210 to P is predicted to be resistant to a KRAS G12C inhibitor. 5 In some embodiments, the KRAS G12C inhibitor is sotorasib and / or adagrasib. In some embodiments, the patient is predicted to be resistant to each of sotorasib and adagrasib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not sotorasib or adagrasib. 10 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including BRAF V480 to A is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have 15 cancer further comprises selecting the patient for treatment with a therapy that is not dabrafenib+cetuximab. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including EGFR D46 to N is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, and a MAP2K1 / 2 inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some 20 embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not sotorasib or adagrasib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including EGFR F436 to S or L and / or S437 to P and / or FS to PP at positions 436-437 is predicted 25 to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not dabrafenib+cetuximab, such as e.g. trametinib. 30 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including EGFR S720 to F is predicted to be resistant to an EGFR TK inhibitor. In some embodiments, the EGFR TK inhibitor is osimertinib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment 35 with a therapy that is not osimertinib, such as e.g. gefitinib. 7 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including EGFR D1006 to N and / or E1005 to K and / or E1004 to K is predicted to be resistant to an EGFR TK inhibitor. In some embodiments, the EGFR TK inhibitor is osimertinib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment 5 for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not osimertinib, such as e.g. gefitinib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including EGFR D1012 to N, D1014 to N, and / or E1015 to K is predicted to be resistant to an EGFR TK inhibitor. In some embodiments, the EGFR TK inhibitor is osimertinib. In some embodiments, a 10 method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not osimertinib, such as e.g. gefitinib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of EGFR L1038 leading to a new splice site resulting in truncation of the EGFR 15 protein after position 1038 is predicted to be resistant to an EGFR TK inhibitor. In some embodiments, the EGFR TK inhibitor is osimertinib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not osimertinib, such as e.g. gefitinib. 20 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of EGFR E1091 leading to a new splice site resulting in truncation of the EGFR protein after position 1091 is predicted to be sensitive to an EGFR TK inhibitor. In some embodiments, the EGFR TK inhibitor is gefitinib and / or osimertinib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been 25 diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that includes gefitinib and / or osimertinib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 Q45 to R and / or Q46 to R is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, a MAP2K1 / 2 inhibitor, and / or a KRAS G12C 30 inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, the G12C inhibitor is adagrasib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib, 35 dabrafenib+cetuximab or adagrasib. 8 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 F53 to L or S is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, a MAP2K1 / 2 inhibitor, and / or a KRAS G12C inhibitor. In some embodiments, the G12C inhibitor is adagrasib. In some embodiments, the G12C inhibitor is sotorasib. In 5 some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib, dabrafenib+cetuximab or adagrasib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein 10 variants including mutation of MAP2K1 Q56 to R and / or K57 to E or R and / or Q58 to R is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, a MAP2K1 / 2 inhibitor, and / or a KRAS G12C inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, the G12C inhibitor is adagrasib and / or sotorasib. In some embodiments, a method for predicting the therapeutic 15 response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib, dabrafenib+cetuximab, sotorasib or adagrasib, or selecting the patient for treatment with an intermittent schedule of one or more of trametinib, dabrafenib+cetuximab, sotorasib or adagrasib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein 20 variants including mutation of MAP2K1 I111 to T and / or Q110 to H and / or I112 to T is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor and / or a MAP2K1 / 2 inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been 25 diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib or dabrafenib+cetuximab. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 L115 to P is predicted to be resistant to a MAP2K1 / 2 inhibitor. In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. some embodiments, a method for predicting the 30 therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib, such as e.g. dabrafenib+cetuximab. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 H119 to R and / or E120 to G is predicted to be resistant to a 35 combination of a BRAF inhibitor and an EGFR inhibitor, a MAP2K1 / 2 inhibitor, and / or a KRAS G12C inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, the G12C inhibitor is 9 adagrasib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib, dabrafenib+cetuximab or adagrasib, or selecting the patient for treatment with an intermittent schedule of 5 one or more of trametinib, dabrafenib+cetuximab or adagrasib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 N122 to S or D and / or E120 to G is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, a MAP2K1 / 2 inhibitor, and / or a KRAS G12C inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. 10 In some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, the G12C inhibitor is adagrasib and / or sotorasib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib, dabrafenib+cetuximab, sotorasib or adagrasib, or selecting the patient for treatment with an 15 intermittent schedule of one or more of trametinib and dabrafenib+cetuximab. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 Y130 to C or H is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, and / or a MAP2K1 / 2 inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some embodiments, the MAP2K1 / 2 20 inhibitor is trametinib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib or dabrafenib+cetuximab, or selecting the patient for treatment with an intermittent schedule of one or more of trametinib and dabrafenib+cetuximab. 25 In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 Y130 to C is predicted to be resistant to a KRAS G12C inhibitor. In some embodiments, the G12C inhibitor is sotorasib and / or adagrasib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient 30 for treatment with a therapy that is not sotorasib and / or adagrasib. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 E203 to K and / or D208 to N is predicted to be resistant to a combination of a BRAF inhibitor and an EGFR inhibitor, and / or a MAP2K1 / 2 inhibitor. In some embodiments, the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab. In some 35 embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been 10 diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib or dabrafenib+cetuximab. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein variants including mutation of MAP2K1 V211 to A is predicted to be resistant to a MAP2K1 / 2 inhibitor. In 5 some embodiments, the MAP2K1 / 2 inhibitor is trametinib. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises selecting the patient for treatment with a therapy that is not trametinib, such as e.g. dabrafenib+cetuximab. In some embodiments, a patient who has been detected as having one or more mitogen pathway protein 10 mutations disclosed herein, such as those of the first aspect, has also been detected as having a BRAF V600E mutation. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises treating the patient with the one or more drugs that the patient has been selected or 15 recommended for treatment with. In some embodiments, a method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer further comprises detecting the presence of the one or more mitogen pathway protein variants in a sample from said patient. 20 Also described herein according to a second aspect is a method of identifying a tumour that is resistant to a cancer treatment, comprising determining in a sample of said tumour the presence or absence of a mitogen pathway protein mutation or gene encoding said protein mutation, whereby the presence of said protein mutation or said gene encoding said protein mutation indicates the tumour is resistant to said cancer treatment, wherein the protein mutation is a E63K / E62K KRAS protein mutation, and the cancer treatment 25 is dabrafenib-cetuximab, or wherein the protein mutation is a K117R / E-D119G KRAS protein mutation, and the cancer treatment is dabrafenib-cetuximab and gefitinib, or wherein the protein mutation is a K499E / R BRAF protein mutation, and the cancer treatment is dabrafenib-cetuximab and trametinib, or wherein the protein mutation is a D1006N / E1005K / E1004K EGFR protein mutation, and the cancer treatment is osimertinib, or wherein the protein mutation is a D1012N / D1014N / E1015K EGFR protein mutation, and the 30 cancer treatment is osimertinib. Also described herein according to a third aspect is a method for identifying a tumour that is sensitised to treatment with trametinib, comprising determining in a sample of said tumour the presence or absence of a protein mutation or gene encoding said protein mutation, whereby the presence of said protein mutation or said gene encoding said protein mutation indicates the tumour is sensitised to treatment with trametinib, 35 wherein the protein mutation is a E63K / E62K KRAS protein mutation, or wherein the protein mutation is a K117R / E-D119G KRAS protein mutation. 11 Also described herein is a method for recommending a treatment for a patient with a tumour, comprising identifying a tumour that is resistant to treatment using the method of the second aspect above, or sensitised to treatment with trametinib according to the third aspect above, wherein the protein mutation is a E63K / E62K KRAS protein mutation, or wherein the protein mutation is a K117R / E-D119G KRAS protein 5 mutation, further comprising recommending treatment with trametinib. Also described herein according to a fourth aspect is a method for identifying a tumour that is addicted to a cancer treatment, comprising determining in a sample of said tumour the presence or absence of a mitogen pathway protein mutation or gene encoding said protein mutation, whereby the presence of said protein mutation or said gene encoding said protein mutation indicates the tumour is addicted to said cancer 10 treatment, wherein the protein mutation is I99T / L98P MAP2K1 and the cancer drug is trametinib, or wherein the protein mutation is S194P MAP2K1 and the cancer drugs are trametinib, dabrafenib-cetuximab, sotorasib and adagrasib, or wherein the protein mutation is Y134H MAP2K2 and the cancer drugs are trametinib, dabrafenib-cetuximab, sotorasib, adagrasib and gefitinib. Also described herein is a method for recommending a treatment for a patient with a tumour, comprising 15 identifying a tumour that is addicted to treatment using the method of the fourth aspect, further comprising recommending a drug holiday. Also described herein is a method of determining the prognosis of a patient, comprising identifying a tumour that is addicted to cancer treatment according to methods of the first or fourth aspects, wherein the protein mutation is I99T / L98P MAP2K1 and the cancer drug is trametinib, or wherein the protein mutation is S194P 20 MAP2K1 and the cancer drugs are trametinib, dabrafenib-cetuximab, sotorasib and adagrasib, or wherein the protein mutation is Y134H MAP2K2 and the cancer drugs are trametinib, dabrafenib-cetuximab, sotorasib, adagrasib and gefitinib, wherein a patient with a tumour identified as being addicted to cancer treatment is assigned a poorer prognosis than a patient with a tumour identified as not being addicted to cancer treatment. 25 In some embodiments, the patient with a tumour identified as being addicted to cancer treatment and assigned a poorer prognosis is selected for a more aggressive treatment regimen than a patient with a tumour identified as not being addicted to cancer treatment. In some embodiments, a more aggressive treatment regimen comprises selecting the patient for surgery where they otherwise may not have been selected; increasing the urgency of a patient’s surgery and / or moving the date of a patient’s surgery 30 closer;changing the surgical plan for the patient’s already planned surgery; selecting the patient for radiotherapy and / or chemotherapy where they otherwise may not have been selected; increasing the urgency, frequency, strength / dose, number of sessions / doses of radiotherapy and / or chemotherapy, and / or moving the start date of radiotherapy and / or chemotherapy closer. Also described herein according to a fifth aspect is a method for identifying a tumour that is sensitised to 35 treatment with the EGFR inhibitors gefitinib and osimertinib, comprising determining in a sample of said tumour the presence or absence of an E1091 or L1038 EGFR protein mutation or gene encoding said 12 protein mutation, whereby the presence of said protein mutation or said gene encoding said protein mutation indicates the tumour is sensitised to treatment with gefitinib and / or osimertinib. In some embodiments, said mutation results in a splice variant in the EGFR gene. In some embodiments, said mutation results in premature termination of EGFR translation. In some embodiments, said mutation results 5 in a C-terminal truncation of EGFR after E1091 or L1038. Also described herein is a method for recommending a treatment for a patient with a tumour, comprising identifying a tumour that is sensitised to treatment with gefitinib and / or osimertinib using the method of the first or fifth aspects, further comprising recommending treatment with gefitinib and / or osimertinib. Also described herein according to a sixth aspect is a MAP2K1 / 2 inhibitor, optionally trametinib, for use in 10 a method of treatment of cancer in a human patient, the method comprising: a) identifying a tumour that is susceptible to treatment with trametinib according to a method of the first or third aspect, and b) treating the patient whose tumour has been identified as susceptible to treatment with the MAP2K1 / 2 inhibitor in step (a) with the MAP2K1 / 2 inhibitor, optionally using an intermittent therapeutic scheme. Also described herein according to a seventh aspect is an EGFR inhibitor, optionally selected from the 15 group of gefitinib and osimertinib, for use in a method of treatment of cancer in a human patient, the method comprising: a) identifying a tumour that is susceptible to said EGFR inhibitor according to a method of the first or fifth aspect, and b) treating the patient whose tumour has been identified as susceptible to said EGFR inhibitor in step (a) with said EGFR inhibitor, optionally using an intermittent therapeutic scheme. Also described herein according to an eighth aspect is a BRAF inhibitor, optionally dabrafenib, and / or an 20 EGFR inhibitor, optionally cetuximab, for use in a method of treatment of cancer in a human patient, the method comprising: a) identifying a tumour that is susceptible to said BRAF inhibitor and / or said EGFR inhibitor according to a method of the first aspect, and b) treating the patient whose tumour has been identified as susceptible to said BRAF inhibitor and / or said EGFR inhibitor in step (a) with said BRAF inhibitor and / or said EGFR inhibitor, optionally using an intermittent therapeutic scheme. 25 Also described herein according to a ninth aspect is a KRAS inhibitor, optionally a KRAS G12C inhibitor or a pan-KRAS inhibitor, optionally sotorasib or adagrasib, for use in a method of treatment of cancer in a human patient, the method comprising: a). identifying a tumour that is susceptible to said KRAS inhibitor according to a method of the first aspect, and b) treating the patient whose tumour has been identified as susceptible to said KRAS inhibitor in step (a) with said KRAS inhibitor, optionally using an intermittent 30 therapeutic scheme. Also described herein according to a tenth aspect is a PI3K inhibitor, optionally a pan-PI3K inhibitor, optionally pictilisib, for use in a method of treatment of cancer in a human patient, the method comprising: a) identifying a tumour that is susceptible to said PI3K inhibitor according to a method of the first aspect, and b) treating the patient whose tumour has been identified as susceptible to said PI3K inhibitor in step 35 (a) with said PI3K inhibitor, optionally using an intermittent therapeutic scheme. 13 Also described herein according to an eleventh aspect is the use of an anticancer agent in the preparation of a medicament for the treatment of a tumour having a mutation in a mitogen pathway protein, wherein a) the anticancer agent is a MAP2K1 / 2 inhibitor, optionally trametinib, and the mutation includes KRAS E62 and / or E63 to K, KRAS K117 to E, R and / or D119 to G, MAP2K2 Y134 to H, and / or MAP2K1 S194 to P; b) 5 the anticancer agent is an EGFR inhibitor, optionally selected from the group of gefitinib and osimertinib, and the mutation includes MAP2K2 Y134 to H, MAP2K1 S194 to P, MAP2K1 L98 to P and / or I99 to T, and / or a variant causing a C-terminal truncation of EGFR after amino acid E1091 or L1038; c) the anticancer agent is a BRAF inhibitor, optionally dabrafenib, and / or an EGFR inhibitor, optionally cetuximab, and the mutation includes MAP2K2 Y134 to H, MAP2K1 S194 to P, MAP2K1 L98 to P and / or I99 to T; d) 10 the anticancer agent is a KRAS inhibitor, optionally a KRAS G12C inhibitor or a pan-KRAS inhibitor, optionally sotorasib or adagrasib, and the mutation includes KRAS E62 and / or E63 to K, KRAS K117 to E, R and / or D119 to G, MAP2K2 Y134 to H, and / or MAP2K1 S194 to P, MAP2K1 L98 to P and / or I99 to T, and / or BRAF K499 to E or R; e) the anticancer agent is a PI3K inhibitor, optionally a pan-PI3K inhibitor, optionally pictilisib, and the mutation includes KRAS E62 and / or E63 to K, KRAS K117 to E, R and / or D119 15 to G, MAP2K2 Y134 to H, and / or MAP2K1 S194 to P, MAP2K1 L98 to P and / or I99 to T, and / or BRAF K499 to E or R. Also described herein according to a twelfth aspect is a nucleic acid probe capable of specifically hybridising to nucleic acid encoding a mutated mitogen pathway protein or fragment thereof incorporating one or more mitogen pathway protein variants, wherein the variants are selected from: variants that are associated with 20 a C-terminal truncation of EGFR, and in particular a truncation after amino acids E1091 or L1038; EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, 25 Y130H, S194P, E203K, D208N, V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, K499E, K499R; PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A; and AKT1 variants selected from: E17K, G16E. Also described herein according to a thirteenth aspect is a method of identifying a mitogen pathway protein mutation in a sample comprising contacting nucleic acid from said sample with a nucleic acid probe 30 according to the twelfth aspect, and detecting said hybridisation. Also described herein according to a fourteenth aspect is a method of detecting the presence of a truncated variant of EGFR in a sample comprising amplifying from said sample nucleic acid corresponding to EGFR or a part thereof, and comparing the electrophoretic mobility of the amplified nucleic acid to the electrophoretic mobility of corresponding wild-type gene or fragment thereof. In some embodiments, the 35 truncated variant has a C-terminal truncation after E1091, optionally after L1038. Also described herein according to a fifteenth aspect is a method of screening for compounds that overcome resistance of a cell that incorporates a mitogen pathway protein variant that is associated with 14 response to one or more drugs, wherein the mitogen pathway protein variants are selected from: EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, 5 K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, K499E, K499R; PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A; and AKT1 variants selected from: E17K, G16E, wherein the presence of the one or more mitogen pathway protein variants are indicative of resistance to 10 one or more of: MAP2K1 / 2 inhibitors, BRAF inhibitors, EGFR inhibitors, PI3K inhibitors, and KRAS inhibitors; and wherein the method comprises: contacting a cell that expresses said one or more mitogen pathway protein variants with one or more candidate compounds or compositions and comparing the effect of the one or more candidate compounds or compositions on said cells with one or more control conditions, and / or contacting said one or more mitogen pathway protein variants with one or more candidate 15 compounds or compositions and comparing the effect of the one or more candidate compounds or compositions on said mitogen pathway protein variant(s) with one or more control conditions, and determining the effect of said one or more candidate compounds or compositions on activity of the mitogen pathway protein variants. The invention includes the combination of the aspects and preferred features described except where 20 such a combination is clearly impermissible or expressly avoided. Brief Description of the Drawings Embodiments and experiments illustrating the principles of the disclosure will now be discussed with reference to the accompanying figures in which: Figure 1a-d shows results demonstrating that base editors map functional domains in driving oncogenes.25 See also Fig.16. Fig. 1a shows an overview of base editing screens to identify drug resistance variants in cancer cell models. Fig.1b shows results of base editor screens in HT-29 cells across 11 cancer genes showing depletion of gRNAs targeting essential genes demonstrating base editing activity. Unpaired, two-tailed Student’s 30 t-test comparing non-targeting gRNAs to gRNAs targeting essential gene splice sites. Boxplots represent the median, inter-quartile range (IQR) and whiskers are the lowest and highest values within 1.5 x IQR. Fig. 1c shows a comparison of gRNA z-scores from base editor screens in PC9 (EGFR-mutant, MYC-dependent) and HT-29 (BRAF-mutant, MYC-dependent) reveal shared and disparate genetic 35 dependencies on driving oncogenes. 15 Fig. 1d shows results of base editing mutagenesis screens of the driving oncogene, BRAF, in HT-29 cells reveal functional protein domains, sites of post-translational modification and driver variants. Data are the average of two independent experiments. Figure 2a-f show results illustrating the classification of variants modulating drug sensitivity cluster into four5 functional classes. Predicted amino acid editing consequences are labelled for drug resistance screens and genotyped edits are shown in Fig. 2d, 2e and 2f. Data in Fig. 2e and 2f are the average of two independent experiments performed on separate days, or representative of two independent experiments. See also Fig.17 and 18. Fig. 2a shows a comparison of gRNA z-scores for the control treated arm vs plasmid library, and the10 drug-treated arm vs plasmid library. Variants conferring resistance to the MEK inhibitor, trametinib, profiled with CBE and ABE base editors in HT-29 CRC cells. Fig. 2b shows a comparison gRNA z-scores for the control treated arm vs plasmid library, and the drug-treated arm vs plasmid library. Variants conferring resistance to the combination of BRAF and EGFR inhibitors, dabrafenib and cetuximab, profiled with CBE and ABE base editors in HT-29 CRC cells.15 Fig. 2c shows a crystal structure of the complex of EGFR and cetuximab (PDB: 1yy9) (Li 2005), andMEK1 and trametinib (PDB: 7jur) (Khan 2020), highlighting canonical drug resistance variants discovered in base editor screens predicted to disrupt drug binding. Fig.2d shows the cell growth of base edited HT-29 cells harbouring canonical and drug addiction drug resistance variants. Cells were untreated or treated with trametinib (3 nM) or the combination of 20 dabrafenib (80 nM) and cetuximab (1 µg / ml), and cell proliferation was monitored using an incucyte. Data represent the mean ± SD of biological triplicates and are representative of two independent experiments. Fig. 2e shows results of Western blotting of wild-type HT-29 ABE cells and cells harbouring drugresistance mutations activating the MAPK signalling pathway. Cells were treated with the combination 25 of dabrafenib (80 nM) and cetuximab (1 µg / ml) or DMSO as a control for 24 h before analysis. Fig.2f shows results of ß-galactosidase staining for senescent cells. ß-galactosidase positive senescent foci (blue) are indicated with arrows. Cells were treated with the combination of dabrafenib (80 nM) and cetuximab (1 µg / ml) or DMSO as a control for 48 h before analysis. Representative images are shown for the drug addiction variant MAP2K1 Y130C / H.30 Figure 3a-c show results illustrating driver variants conferring drug resistance. Predicted amino acid editingconsequences are labelled for drug resistance variants. Data are the average of two independent screens performed on separate days. Fig.3a shows a comparison of total COSMIC mutation counts for amino acid positions compared to the z-score of gRNAs tiling across AKT1 and PIK3CA in HT-29 cells. The results show that base editing35 screens reveal clinically apparent hotspot mutations in oncogenes. 16 Fig. 3b shows a comparison of gRNA z-scores for the control treated arm vs plasmid library, and the drug-treated arm vs plasmid library. Drug resistance variants to the PI3K inhibitor, pictilisib, profiled with CBE and ABE base editors in HT-29 CRC cells. Fig.3c shows a comparison of gRNA z-scores for the sotorasib treated arm vs plasmid library, and the 5 adagrasib treated arm vs plasmid library. Drug resistance variants to the KRASG12Cinhibitors, sotorasib and adagrasib, profiled with CBE and ABE base editors in H23 lung cancer cells. Figure 4A-B show results demonstrating that base editing screens map functional domains in drivingoncogenes. Replicate correlation for CBE and ABE screens across three cancer cell models; HT-29, H23, PC9. Pearson correlation coefficient values (R) between independent replicate screens are shown. Low 10 correlation was observed for replicates of PC9 screens with gefitinib, which may relate to a high degree of enrichment of resistant, EGFR T790M base edit harbouring cells. Fig.4A shows results for HT-29 (all conditions) and H23 (CBE control, CBE sotorasib, CBE adagrasib). Fig.4B shows results for H23 (ABE control, ABE sotorasib, ABE adagrasib) and PC9 (all conditions.) Figure 5a-d show results of experiments investigating drug resistance and drug-sensitising variants in15 EGFR. Data are normalised to day 0 ratios and represent the mean ± SD of biological triplicates. Unpaired, two-tailed Student’s t-test comparing to the EGFR C797C synonymous variant control; *p-value <0.0005. Predicted amino acid editing consequences are labelled for drug resistance variants screens. See also Fig. 20. Fig. 5a shows a comparison of gRNA z-scores for the control treated arm vs plasmid library, and the 20 drug-treated arm vs plasmid library. Drug resistance variants to the EGFR inhibitor, gefitinib, profiled with CBE and ABE base editors in PC9 lung cancer cells. Fig. 5b shows a comparison of gRNA z-scores for the control treated arm vs plasmid library, and the drug-treated arm vs plasmid library is shown. Drug resistance variants to the EGFR inhibitor, osimertinib, profiled with CBE and ABE base editors in PC9 lung cancer cells. Screening data are the average of 25 two independent screens performed on separate days. Fig. 5c results of prime editing mutagenesis screens of EGFR in the presence and absence of osimertinib. PC9 MLH1 KO cells were prime edited for 7 days with doxycycline (1 µg / ml) before growth for 10 days in DMSO (control), or osimertinib (75 nM). Data are the z-score for each pegRNA derived from the average of two independent screens performed on separate days. Samples were compared to 30 the plasmid library. Fig. 5d shows results of co-competition flow cytometry assays in PC9 MLH1 KO cells comparing the growth of NT gRNA GFP cells to epegRNA BFP cells harbouring different EGFR variants in the presence and absence of osimertinib (75 nM) for 5 days. Figure 6a-d show results of experiments demonstrating that EGFR C-terminal truncating variants sensitise35 to EGFR inhibitors. NT, non-targeting gRNA. See also Fig.21. Fig. 6a shows drug response in PC9 lung cancer cells to EGFR inhibitors confirming sensitisation of EGFR C-terminal truncating mutants to gefitinib and osimertinib. Data represent the mean ± SEM of two 17 independent experiments, each performed in biological triplicate. Two-way ANOVA comparing to parental response; ***p-value <0.0001. Fig.6b shows that a drug-sensitising base edit in EGFR causes loss of a splice donor site. The EGFR RNA splice variants are shown by migration of PCR products from cDNA. The larger PCR product in 5 the mutant samples is due to retention of a short region of a downstream intron sequence after exon 7, where an alternative splice donor is used. EGFR protein after residue 1091 is not translated due to a frame-shift leading to a stop codon. Fig.6c shows results of Western blotting of drug-sensitising mutants revealing a C-terminal truncation in EGFR and confirms drug sensitisation. PC9 CBE or ABE control cells (NT gRNA) or cells mutant for 10 EGFR were treated with gefitinib or osimertinib for 24 h before analysis. Data are representative of two independent experiments. Fig.6d shows results of a flow cytometry analysis of EGFR protein surface expression in PC9 cells with wild-type EGFR or base edited EGFR. Data are represented as a histogram or quantified as EGFR- FITC mean fluorescence intensity (MFI), and represent the mean of three independent experiments ± 15 SD. Unpaired, two-tailed Student’s t-test; ***p-value <0.001, **p-value <0.01, *p-value <0.05. Figure 7a-g show results desmonstrating that perturb-seq functionally defines drug resistant cell states.NT; non-targeting gRNA. Control gRNAs are those that did not confer drug resistance in proliferation screens. See also Fig.22-24. Fig.7a shows a schematic of perturb-seq screening to investigate the transcriptomic effects of variants 20 conferring resistance to dabrafenib and cetuximab in HT-29 cells using base editing. Fig. 7b shows UMAPs coloured by variant class and normalised energy distances (ed) between NT gRNA cells and drug resistant cells in ABE HT-29 cells treated with the combination of dabrafenib (80 nM) and cetuximab (1 µg / ml) for 16 h. Fig. 7c shows cell cycle phase occupancy differences between cells with drug resistance conferring 25 gRNAs and control gRNAs in the ABE perturb-seq dataset. Fig. 7d shows a heatmap and hierarchical clustering of PROGENy pathway activity scores for each gRNA in the ABE dataset. Fig. 7e shows a density plot of differences in PROGENy pathway scores between the variant groupsfor the ABE dataset. 30 Fig.7f shows a volcano plot of differentially expressed genes between NT gRNA control cells and cells with the PI3K p110ɑ driver variant. FDR, false discovery rate. Significantly downregulated transcriptsare in red (including B2M and HLA-A), and upregulated transcripts are in blue. Fig.7g shows a boxplot of progression-free survival (PFS) outcome score for each variant class, derived from CRC patients treated with BRAF, MEK and PD-1 inhibitor combination therapy (Tian 2023). CBE 35 and ABE perturb-seq scores are shown. ***P-value < 0.01, Wilcoxon rank-sum test. Figures 8A-B show a drug resistance variant map indicating potential second-line therapies. See also Fig.25, Fig.16 and Tables 5A-D. 18 Fig.8a shows base editing efficiency and precision mapped across 45 endogenous loci in HT-29 CBE and ABE cells. Average VAFs for exact edits for hit gRNAs are shown for each variant from amplicon sequencing data that were absent in unedited samples. Dashed lines represent the predicted base editing activity window. Data represent the mean of two independent experiments performed on 5 separate days. VAF, variant allele frequency. Fig. 8b shows a variant function map for variants modulating drug sensitivity in cancer. Potential alternative treatments tested in this study are highlighted. Genotypes are from next-generation sequencing or Sanger sequencing of hits from base editing screens. Figure 9a-c show average z-scores for selected validated variants.10 Fig.9a shows a table of average z-scores of two replicates for each cancer drug using base editing. Fig.9b shows a table of average z-scores of two replicates for each cancer drug using prime editing. Fig.9c shows a table of average z-scores of two replicates for each cancer drug using prime editing. Figure 10 shows a table of differential energy of iBAR groups from non-targeting (NT) control gRNAs (ABE).Table shows iBAR groups with each gRNA, for selected variants. Variants are classed as in the analysis of 15 pooled screens / perturb-seq analysis. Only variants with non-control variant class are shown. Table shows number of cells with gRNA called, number of different iBAR barcodes with the gRNA, gene expression based cluster of the gRNA, energy distance from non-targeting control, diffusion score, PFS outcome score being a measure of similarity of drug response to that in patients with PFS > 6 months (Tian 2023), proportion of iBAR clones in the cluster with the higher diffusion scores (and therefore more advanced to 20 full drug addiction) for gRNAs with a bimodal distribution of energy distances, and binary sign of transcriptional impact, indicating whether or not there is at least one gene or pathway differentially expressed compared to the non-targeting control with a p-value of less than 10-6. Figure 11 shows a table of differential energy of iBAR groups from non-targeting (NT) control gRNAs(CBE). Table shows iBAR groups with each gRNA, for variants of interest. Variants are classed as in the 25 analysis of pooled screens / perturb-seq analysis. Only variants with non-control variant class are shown. Table shows number of cells with gRNA called, number of different iBAR barcodes with the gRNA, gene expression based cluster of the gRNA, energy distance from non-targeting control, diffusion score, PFS outcome score being a measure of similarity of drug response to that in patients with PFS > 6 months (Tian 2023), proportion of iBAR clones in the cluster with the higher diffusion scores (and therefore more 30 advanced to full drug addiction) for gRNAs with a bimodal distribution of energy distances, and binary sign of transcriptional impact, indicating whether or not there is at least one gene or pathway differentially expressed compared to the non-targeting control with a p-value of less than 10^-6. Figure 12 shows a table of results of PROGENy analysis (ABE). Table shows gRNA, the pathway that isdifferentially expressed compared to non-targeting control, the uncorrected p-value of Wilcoxon rank-sum 35 test (p.value), the Benjamini-Bogomolov corrected p-value (FDR), and the log2-fold change compared to non-targeting control (computed using findMarkers function from scran Bioconductor package (Lun 2016)) (LFC). 19 Figure 13 shows a table of results of PROGENy analysis (CBE). Table shows gRNA, the pathway that isdifferentially expressed compared to non-targeting control, the uncorrected p-value of Wilcoxon rank-sum test (p.value), the Benjamini-Bogomolov corrected p-value (FDR), and the log2-fold change compared to non-targeting control (computed using findMarkers function from scran Bioconductor package (Lun 2016)) 5 (LFC). Figure 14 shows results of a MAYA / MSigDB Hallmark analysis (ABE). Pathway differential analysis wasrepeated using MAYA (Landais 2023) with the MSigDB Hallmark pathways (Liberzon 2015) as input gene lists. The table shows pathways for which MAYA calculated bimodality in the PCA data. The inventors then calculated the uncorrected p-value of Wilcoxon rank-sum test (p.value), the Benjamini-Bogomolov 10 corrected p-value (FDR), and the log2-fold change compared to non-targeting control (computed using findMarkers function from scran Bioconductor package (Lun 2016)) (LFC), as for the PROGENy analysis. Figure 15 shows results of a MAYA / MSigDB Hallmark analysis (CBE). Pathway differential analysis wasrepeated using MAYA (Landais 2023) with the MSigDB Hallmark pathways (Liberzon 2015) as input gene lists. The table shows pathways for which MAYA calculated bimodality in the PCA data. The inventors then 15 calculated the uncorrected p-value of Wilcoxon rank-sum test (p.value), the Benjamini-Bogomolov corrected p-value (FDR), and the log2-fold change compared to non-targeting control (computed using findMarkers function from scran Bioconductor package (Lun 2016)) (LFC), as for the PROGENy analysis. Figure 16 shows genotyping results including VAF from NGS amplicon sequencing of endogenous base edits for selected hit gRNAs. VAF, variant allele frequency. Full table provided as Table 5D. “Edit”, actual 20 edit determined by amplicon sequencing. “Protein change”, predicted protein change from the base editing gRNA. Figure 17a-d show results demonstrating that variants modulating drug sensitivity cluster into four functional classes. Fig. 17a shows results of base editor screens in H23, PC9 and MHH-ES-1 cancer cells targeting 11 25 cancer genes, showing depletion of gRNAs targeting essential genes demonstrating base editing activity. Unpaired, two-tailed Student's t-test comparing non-targeting gRNAs to gRNAs targeting essential gene splice sites. Boxplots represent the median and interquartile range (IQR), and whiskers represent the lowest and highest values within 1.5 x the IQR. Fig.17b shows the number of off-target sites plotted against the z-score for base editing gRNAs. A high 30 number of off-targets for a small number of KRAS-targeting gRNAs is associated with severe gRNA depletion. These were filtered out of downstream analysis. Fig. 17c shows results from a previously reported whole-genome CRISPR-Cas9 KO screen in HT-29 cells in the presence of dabrafenib (0.1 µM) across three time-points. Volcano plot showing EGFR KO as the top sensitising hit. Data are the average of two independent screens. 35 Fig. 17d shows the output of TCGA oncoprint (pan-cancer cohort, n = 526) of colorectal adenocarcinomas with alterations in KRAS and BRAF. Mutual exclusivity p-value <0.001. 20 Figure 18a-b shows results of validation of cancer drug addiction variant phenotypes.Fig.18a shows results of Western blotting of drug resistance variants from base editing screens in HT- 29 cells conferring resistance to dabrafenib and cetuximab combination therapy. HT-29 cells harbouring the indicated variants were treated with dabrafenib (80 nM) and cetuximab (1 µg / ml) or DMSO (control) 5 for 24 h before analysis. Data are representative of two independent experiments (see also Fig.2). Fig. 18b shows microscopy images of ß-galactosidase assays performed to measure the induction of senescence. HT-29 cells harbouring the indicated variants were treated with dabrafenib (80 nM) and cetuximab (1 µg / ml) or DMSO (control) for 48 h before analysis. Representative images from two independent experiments. Scale bar indicates 500 µm. Genotyped variants are shown. 10 Figure 19 shows results for drug-sensitising variants. Variants modulating sensitivity to PARP1 / 2 inhibitors olaparib or niraparib in MHH-ES-1 cells in CBE or ABE screens. Comparison of gRNA z-scores for the drug-treated arm vs plasmid library against the z-scores from the untreated control vs the plasmid library. Predicted edited amino acid positions are labelled. Figure 20a-c show results of prime editing screening of EGFR variants.15 Fig. 20a shows that a Western blot for MLH1 verifies KO of MLH1 in PC9 cells. PC9 cells were transfected with a Cas9-GFP plasmid encoding an MLH1 gRNA. FACS of GFP positive single cells gave clonal populations, or a pooled population (“pool”). Cells were expanded before analysis by Western blotting. Actin serves as a loading control. Fig. 20b shows results of Sanger sequencing of prime editing of EGFR C797S in PC9 cells. PC9-PE20 MLH1 KO (clone 1 from above), or MLH1 WT PC9-PE cells were infected with a pegRNA encoding the C797S edit, puromycin selected and prime editing was initiated with the addition of doxycycline for 5 days. Control (untreated) or osimertinib selected cells (5 nM) are shown. The EGFR C797 locus was PCR amplified and then analysed with Sanger sequencing. Fig.20c shows replicate correlation between pegRNA z-scores from EGFR prime editing mutagenesis 25 screens performed in PC9 MLH1 KO PE2 cells. Data are from two independent screens performed on different days. Labelled are predicted mutations in EGFR installed by the pegRNAs. Pearson correlation coefficient values (R) between independent replicate screens are shown. pegRNA, prime editing gRNA. Figure 21a-d show results demonstrating that EGFR C-terminal truncating variants sensitise to EGFRinhibitors.30 Fig. 21a shows results of drug titration experiments in PC9 CBE and ABE cells using Cell-titre Glo tomeasure cell proliferation in the presence of EGFR inhibitors (cetuximab, erlotinib, lapatinib), or chemotherapy agents (cisplatin, paclitaxel). Data represent the mean ± SD of two independent experiments performed on separate days, each in biological triplicate. Fig.21b shows results of Sanger sequencing of DNA from WT or base edited PC9 cells harbouring the 35 EGFR-inhibitor sensitising splice variant. CBE editing and ABE editing of a known (GT) splice donor is shown. The position of each gRNA is indicated. 21 Fig.21c shows results of Sanger sequencing cDNA from WT or base edited PC9 cells harbouring the EGFR-inhibitor sensitising splice variant. WT cells display exon-exon splicing as expected, whereas mutant cells display intron retention by utilising an alternative splice donor in the downstream intron. Fig. 21d Gating strategy for flow cytometry analysis of EGFR expression on PC9 cells (FITC). Gating 5 was performed on cells, singlets, viable cells, BFP+ cells (gRNA expression). Figure 22a-e show results of Perturb-seq quality control and pathway analysis. Fig. 22a shows the correlation between large-scale base editing screens (PC9 CBE and ABE) and a small-scale validation base editing screen designed for perturb-seq. Pearson correlation coefficients are shown for gefitinib and osimertinib screens. 10 Fig. 22b shows a density plot of gRNA classes against cell numbers in single-cell sequencing for HT- 29 CBE and ABE experiments after quality control. Cells with gRNAs targeting splice sites in essential genes are depleted, indicating efficient editing. Fig.22c shows a heatmap of scaled expression levels (mean=0, SD=1, average across gRNA) of genes and are differentially expressed for at least one resistance gRNA with an absolute log2-fold change > 15 0.5 at FDR < 0.1 for at least one gRNA when comparing against cells with NT gRNAs in HT-29 CBE or ABE perturb-seq screens. The dendrogram was cut at 4 clusters to show the varying gene expression levels and their association with variant class. Fig. 22d shows UMAPs coloured by variant class and normalised energy distances (ed) between NT gRNA cells and drug resistant cells in CBE HT-29 cells treated with the combination of dabrafenib (80 20 nM) and cetuximab (1 µg / ml) for 16 h. Fig.22e shows a heatmap of scaled expression levels (mean=0, SD=1, average across gRNA) of cell- cycle related genes (GO.0007049) that are differentially expressed for at least one resistance gRNA with absolute log2-fold change > 0.75 and FDR < 0.001 for the HT-29 ABE perturb-seq screen for at least one gRNA. 25 Figure 23a-e. show results demonstrating that Perturb-seq functionally defines drug resistant cell states. Fig. 23a shows results of a differential gene expression analysis using pathways from MAYA or PROGENy for HT-29 CBE and ABE perturb-seq screens. Heatmaps display log-fold changes for a given pathway-gRNA comparison, and statistical significance is denoted with a dot (significance at FDR<0.1). Differential expression of pathway scores compared to NT gRNAs. 30 Fig. 23b shows results of a differential expression at the level of PROGENy pathway scores for drug addiction versus canonical drug resistance. For each gRNA the number of iBARs was selected to avoid biases resulting from an over-representation of individual gRNAs. Fig. 23c shows a comparison of z-scores from proliferation read-out base editing screens to energy distance scores derived from perturb-seq screens. Variant classes based on the HT-29 proliferation 35 screens in dabrafenib and cetuximab are indicated. Intermediate variants discussed in the text are labelled. 22 Fig.23d shows diffusion scores illustrating progressive levels of mutational impact for the CBE and ABE data set, with drug addiction variants having the highest scores and a range of different impact levels across the gRNAs conferring drug resistance. The intermediate variants KRAS E62K / E63K and KRAS K117R / E / D119G are highlighted. 5 Fig.23e shows volcano plots of significantly differentially expressed genes (vs NT control gRNA cells) from representative drug resistance gRNAs. B2M is downregulated by both variants. Significant down- and upregulation at FDR<0.1 (corrected across targets and across transcriptome for each target) are indicated in blue and red respectively. Figure 24a-e show results demonstrating that Perturb-seq functionally defines drug resistant cell states. 10 Fig.24a show results of a flow cytometry assessment of B2M and HLA-A,B,C expression in HT-29 ABE cells harbouring drug addiction variants. Data represent the mean ± SD of biological triplicates. IFN- gamma treatment serves as a positive control (48 h, 400 U / ml). ****P-value <0.0001; ***P-value <0.001; **P-value <0.01; *P-value <0.05; unpaired, two-tailed Student's t-test comparing to non-targeting gRNA (NT) condition. Genotyped variants are shown. 15 Fig. 24b shows results of a flow cytometry assessment of B2M and HLA-A,B,C expression in CRC-9 ABE tumour organoid cells harbouring drug addiction variants. Cells were treated with DMSO (control) or the MEK inhibitor trametinib (25 nM) for 48 h before analysis. Data represent the mean ± SD of two independent experiments, each with two-three replicates. IFN-g treatment serves as a positive control (48 h, 400 U / ml). ****P-value <0.0001; ***P-value <0.001; **P-value <0.01; *P-value <0.05; unpaired, 20 two-tailed Student's t-test comparing non-targeting gRNA (NT) condition. Genotyped variants are shown. Fig.24c shows a representative flow cytometry gating used for CRC-9 tumour organoids to assess HLA and B2M cell surface protein expression. Single, live cells with mApple (ABE) and BFP (gRNA) expression were gated for analysis. 25 Fig.24d shows results of a co-competition flow cytometry assays of WT (GFP – NT gRNA expressing cells) and drug resistant CRC-9 tumour organoids (BFP – gRNA expressing) at 72 h. Data represent the mean ± SD of biological triplicates. ****P-value <0.0001; ***P-value <0.001; **P-value <0.01; *P- value <0.05; unpaired, two-tailed Student's t-test comparing to non-targeting gRNA (NT) condition. Genotyped variants are shown. 30 Fig. 24e shows results of a co-culture assay of primary, autologous, anti-tumour T cells with CRC-9 tumour organoids harbouring different drug addiction variants. Cancer cells were pre-treated with the MEK inhibitor trametinib (25 nM) for 48 h before washing and plating the co-culture assay plate. Flow cytometry assessment of absolute cell numbers (measured by counting beads) following 72 h co-culture. Data are expressed as the percentage of live cells remaining as compared to the relevant condition in 35 the absence of T cells and represent the mean ± SD of biological triplicates. *P-value <0.05; unpaired, two-tailed Student's t-test. Genotyped variants are shown. 23 Figure 25A, B shows results of next-generation sequencing of base edits across 45 variants modulatingdrug sensitivity. Fig. 25A shows results for BRAF, EGFR, KRAS, MAP2K1, MAP2K2 and PIK3CA. Fig. 25B shows results for AKT1, EGFR and KRAS. Editing efficiency and precision of CBE and ABE base editors are shown by amplicon sequencing of endogenous DNA loci. Base editing was performed by 5 doxycycline-induced expression of ABE (top panel) or CBE (bottom panel) for three days. Rare transversion mutations and their sequence context within the gRNA are highlighted by a red box. VAF, variant allele frequency from amplicon sequencing and represent the mean of two independent experiments. Detailed Description Aspects and embodiments of the present disclosure will now be discussed with reference to the 10 accompanying figures. Further aspects and embodiments will be apparent to those skilled in the art. All documents mentioned in this text are incorporated herein by reference. The present disclosure relates to variants of mitogen pathway proteins that are indicative of response to one or more cancer treatments, and related methods. The present disclosure relates to methods of identifying mitogen pathway protein variants in tumour samples resulting in the resistance, addiction, or 15 sensitising of the tumour to one or more cancer treatments. The wording “indicative of response to one or more inhibitors” encompasses response to therapeutic regimens that include the one or more inhibitors. As used herein, the term “resistance”, also referred to as “drug resistance”, “multiple drug resistance”, “cancer drug resistance”, “canonical drug resistance”, “antineoplastic resistance”, or “chemotherapy 20 resistance”, refers to the resistance of cancerous cells to a particular drug or class of drugs, or the ability of cancer cells to survive and grow despite the presence of a particular drug or class of drugs. In the context of a patient with cancer or tumour, the term “resistance” refers to a lack of response or a lower response of the patient’s cancer or tumour to a drug or class of drugs. In some embodiments, the cancer cells, cancer or tumour is resistant to multiple drugs, resulting in multiple drug resistance. In some embodiments, 25 resistance is the result of inherent genetic characteristics in the cancer cells. In some embodiments, resistance is acquired after drug exposure. For example, tumour heterogeneity, wherein tumours are made up of different populations of cancer cells that are morphologically, phenotypically, genetically and / or functionally different, can result in the effective selection of certain populations of cancer cells which possess characteristics that confer drug resistance, when the tumour is treated with said drug. 30 According to the present disclosure, resistance or sensitivity to a drug or class of drugs is associated with specific mutations (also referred to as “variants”) present in the cancerous cells or a proportion thereof in a tumour or cancer. The term “variant” (also referred to herein as “mutant”) refers to a protein that differs from the wild type protein present in normal / healthy cell, to the particular mutation or change in protein sequence that differs between the variant protein and the wild type / normal protein, and / or to the underlying 35 genetic sequence change that caused the change in the protein. Thus, mutants described herein are all associated with a different protein compared to the wild type protein (i.e. no silent mutation). A variant may 24 be a single or multiple amino acid substitution, a splice variant or a truncated sequence variant. A variant may be associated with a position, which may be expressed as an amino acid number (referring to the position of the amino acid(s) that are varied, in the reference protein sequence) or as a position number in the gene sequence encoding the protein. Unless indicated otherwise, all variant positions are indicated by 5 reference to the amino acid sequence of the corresponding reference protein sequence. Variants associated with drug resistance may be classified in different classes depending on whether cells that have the variant have a proliferation advantage in the presence of the drug and / or in the absence of the drug. In some embodiments, a variant is classified as a “drug resistance variant” (also referred to as “canonical drug resistance variant”) when cancer cells harbouring this variant show a proliferation 10 advantage in the presence of one or more cancer drugs or classes of drugs compared to cancer cells that do not have the variant, but do not show the proliferation advantage in the absence of the drugs. In some embodiments, a variant is classified as a “driver”, “driver variant” or “driver mutation” when cells that have the variant show a proliferation advantage in the presence of one or more cancer drugs or classes of drugs, and in the absence of drugs or classes of drugs, compared to cancer cells that do not have the variant. In 15 some embodiments, the presence of a driver mutation results in clonal expansion and / or an increased rate of proliferation relative to wild-type cells. In some embodiments, a variant is classified as an “addiction” variant, also referred to as “drug addiction”, “cancer drug addiction”, when cells that have the variant show a proliferation advantage in the presence of one or more cancer drugs or classes of drugs, but show a proliferation disadvantage in the absence of 20 the one or more cancer drugs or classes of drugs, compared to cancer cells that do not have the variant. As used herein, the term “sensitising” in relation to a variant refers to the presence of a deleterious effect of a drug or class of drugs on proliferation of cancer cells that have the variant. The deleterious effect may be larger than for cancer cells that do not have the variant. Thus, cancer cells with a sensitising variant may have a proliferation disadvantage in the presence of one or more cancer drugs or classes of drugs, 25 compared to cancer cells that do not have the variant. This may also be seen as a decrease in cancer drug resistance. A proliferation advantage or disadvantage can be assessed using a cell-based assay, comparing proliferation of cancer cells including the variant to proliferation of cancer cells that do not include the variant (control). The control may be assumed to be genetically identical to the cancer cells including the variant 30 other than in relation to the variant genomic position. Cancers comprising cancer cells that have one or more drug resistance variants may be selected for treatment with one or more therapies (including e.g. treatment with one or more drugs) that do not include or are not limited to the one or more drugs or classes of drugs that the resistance variants is / are associated with. This may be particularly recommended for cancers comprising cancer cells that have one or more 35 drug resistance variants that are canonical drug resistance variants or driver variants. Cancers comprising cancer cells that have one or more drug resistance variants that are drug addiction variants may be selected 25 for treatment with one or more drugs classes of drugs that the resistance variants is / are associated with, using an intermittent treatment regimen, also referred to as “drug holiday”. As used herein, the term “drug holiday”, also referred to as a “drug vacation”, “medication vacation”, “structured treatment interruption”, “intermittent treatment”, “tolerance break”, “treatment break”, 5 “medication holiday”, “medication break”, or “strategic treatment interruption”, refers to a patient stopping taking a medication(s) for a period of time. In certain circumstances, administration of a treatment to a patient may be advantageously interrupted for a period of time. The interruption can comprise reducing the dose or frequency of the medication, or ceasing administration entirely. In some embodiments, the drug holiday may be a reduction in the drug (e.g. below the therapeutically effective amount for a certain interval 10 of time). In other embodiments, administration of the drug is stopped for an interval of time before being started again, at the same or different dosage regiment. In some embodiments, the duration of the drug holiday may be at least twice that of the relevant dosing interval, at least 3 times, at least 4 times, at least 5 times, at least 10 times, or at least 20 times that of the relevant dosing interval or mean thereof. In some embodiments, the duration of the drug holiday may be at least one day, at least two days, at least three 15 days, at least a week, at least 2 weeks, at least 4 weeks, at least a month, at least 2 months, at least 3 months, at least 6 months, or more. Cancers comprising cancer cells that have one or more drug sensitising variants may be referred to as “sensitised” or “responsive” to the one or more drugs ot classes or drugs that the sensitising variants is / are associated with. such patients may be selected for treatment with one or more therapies (including 20 treatment with one or more drugs) that include the one or more drugs or classes of drugs that the sensitising variants is / are associated with. Treatment strategies, treatment selection and treatment recommendation can include identification, selection or recommendation (in each case encompassing positive selection or exclusion) of a plurality of drugs each associated with a different variant. For example, a patient may be selected for treatment with a 25 plurality of drugs comprising a drug associated with a sensitising variant present in cancer cells in the patient (positive selection), and a drug that is not a drug associated with a resistance variant present in cancer cells in the patient (exclusion). As another example, a patient may be selected for treatment with a plurality of drugs comprising a drug associated with a drug addiction variant present in cancer cells in the patient (positive selection) administered using an intermittent scheme (drug holiday), and a drug that is not 30 a drug associated with a resistance variant present in cancer cells in the patient (exclusion). Any combinations of variants and associated therapeutic recommendation, selection or treatment may be present in a patient. The term “variant allele fraction” (VAF) refers to the proportion of a population of copies of a genetic sequence encoding a protein, in a population of cells, that have a particular variant. The phenotypic 35 consequences of a variant as described herein may depend on the variant allele fraction with which the variant is present. The variant allele fraction of a variant in a population of cells (e.g. a population of tumour 26 cells) may represent a complex combination of the number of cells that have the variant on any copies of the gene encoding the protein, and the number of copies of the gene encoding the protein in the cells. The term “EGFR” refers to the human Epidermal growth factor receptor (also known as ERBB, ERBB1, HER1) or a homologue thereof. The protein sequence of human EGFR is available under Uniprot accession 5 number P00533 and is incorporated herein in its entirety. The genomic sequence of human EGFR is available under NCBI GeneID 1956, and is incorporated herein in its entirety. The present disclosure provides a plurality of variants in the EGFR protein sequence that are associated with altered response to one or more anti-cancer drugs. These include variants that are associated with a C-terminal truncation of EGFR, and in particular a truncation after amino acids E1091 or L1038. Amino acids E1091 and L1038 are 10 highlighted in bold in the sequence of human EGFR reproduced below: >sp|P00533|EGFR_HUMAN Epidermal growth factor receptor OS=Homo sapiens OX=9606 GN=EGFR PE=1 SV=2 MRPSGTAGAALLALLAALCPASRALEEKKVCQGTSNKLTQLGTFEDHFLSLQRMFNNCEV VLGNLEITYVQRNYDLSFLKTIQEVAGYVLIALNTVERIPLENLQIIRGNMYYENSYALA 15 VLSNYDANKTGLKELPMRNLQEILHGAVRFSNNPALCNVESIQWRDIVSSDFLSNMSMDF QNHLGSCQKCDPSCPNGSCWGAGEENCQKLTKIICAQQCSGRCRGKSPSDCCHNQCAAGC TGPRESDCLVCRKFRDEATCKDTCPPLMLYNPTTYQMDVNPEGKYSFGATCVKKCPRNYV VTDHGSCVRACGADSYEMEEDGVRKCKKCEGPCRKVCNGIGIGEFKDSLSINATNIKHFK NCTSISGDLHILPVAFRGDSFTHTPPLDPQELDILKTVKEITGFLLIQAWPENRTDLHAF 20 ENLEIIRGRTKQHGQFSLAVVSLNITSLGLRSLKEISDGDVIISGNKNLCYANTINWKKL FGTSGQKTKIISNRGENSCKATGQVCHALCSPEGCWGPEPRDCVSCRNVSRGRECVDKCN LLEGEPREFVENSECIQCHPECLPQAMNITCTGRGPDNCIQCAHYIDGPHCVKTCPAGVM GENNTLVWKYADAGHVCHLCHPNCTYGCTGPGLEGCPTNGPKIPSIATGMVGALLLLLVV ALGIGLFMRRRHIVRKRTLRRLLQERELVEPLTPSGEAPNQALLRILKETEFKKIKVLGS 25 GAFGTVYKGLWIPEGEKVKIPVAIKELREATSPKANKEILDEAYVMASVDNPHVCRLLGI CLTSTVQLITQLMPFGCLLDYVREHKDNIGSQYLLNWCVQIAKGMNYLEDRRLVHRDLAA RNVLVKTPQHVKITDFGLAKLLGAEEKEYHAEGGKVPIKWMALESILHRIYTHQSDVWSY GVTVWELMTFGSKPYDGIPASEISSILEKGERLPQPPICTIDVYMIMVKCWMIDADSRPK FRELIIEFSKMARDPQRYLVIQGDERMHLPSPTDSNFYRALMDEEDMDDVVDADEYLIPQ 30 QGFFSSPSTSRTPLLSSLSATSNNSTVACIDRNGLQSCPIKEDSFLQRYSSDPTGALTED SIDDTFLPVPEYINQSVPKRPAGSVQNPVYHNQPLNPAPSRDPHYQDPHSTAVGNPEYLN TVQPTCVNSTFDSPAHWAQKGSHQISLDNPDYQQDFFPKEAKPNGIFKGSTAENAEYLRV APQSSEFIGA Thus, the present disclosure provides variants of human EGFR or a homologue thereof in which the 35 underlined sequence (after the L in bold or after the E in bold) above is not present. EGFR variants of the disclosure further include variants in which any one or more of the following amino acid substitutions occur: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K. These amino acids are highlighted in bold in the sequence of human EGFR reproduced below: 40 >sp|P00533|EGFR_HUMAN Epidermal growth factor receptor OS=Homo sapiens OX=9606 GN=EGFR PE=1 SV=2 MRPSGTAGAALLALLAALCPASRALEEKKVCQGTSNKLTQLGTFEDHFLSLQRMFNNCEV VLGNLEITYVQRNYDLSFLKTIQEVAGYVLIALNTVERIPLENLQIIRGNMYYENSYALA VLSNYDANKTGLKELPMRNLQEILHGAVRFSNNPALCNVESIQWRDIVSSDFLSNMSMDF 45 QNHLGSCQKCDPSCPNGSCWGAGEENCQKLTKIICAQQCSGRCRGKSPSDCCHNQCAAGC TGPRESDCLVCRKFRDEATCKDTCPPLMLYNPTTYQMDVNPEGKYSFGATCVKKCPRNYV VTDHGSCVRACGADSYEMEEDGVRKCKKCEGPCRKVCNGIGIGEFKDSLSINATNIKHFK NCTSISGDLHILPVAFRGDSFTHTPPLDPQELDILKTVKEITGFLLIQAWPENRTDLHAF ENLEIIRGRTKQHGQFSLAVVSLNITSLGLRSLKEISDGDVIISGNKNLCYANTINWKKL 27 FGTSGQKTKIISNRGENSCKATGQVCHALCSPEGCWGPEPRDCVSCRNVSRGRECVDKCN LLEGEPREFVENSECIQCHPECLPQAMNITCTGRGPDNCIQCAHYIDGPHCVKTCPAGVM GENNTLVWKYADAGHVCHLCHPNCTYGCTGPGLEGCPTNGPKIPSIATGMVGALLLLLVV ALGIGLFMRRRHIVRKRTLRRLLQERELVEPLTPSGEAPNQALLRILKETEFKKIKVLGS 5 GAFGTVYKGLWIPEGEKVKIPVAIKELREATSPKANKEILDEAYVMASVDNPHVCRLLGI CLTSTVQLITQLMPFGCLLDYVREHKDNIGSQYLLNWCVQIAKGMNYLEDRRLVHRDLAA RNVLVKTPQHVKITDFGLAKLLGAEEKEYHAEGGKVPIKWMALESILHRIYTHQSDVWSY GVTVWELMTFGSKPYDGIPASEISSILEKGERLPQPPICTIDVYMIMVKCWMIDADSRPK FRELIIEFSKMARDPQRYLVIQGDERMHLPSPTDSNFYRALMDEEDMDDVVDADEYLIPQ 10 QGFFSSPSTSRTPLLSSLSATSNNSTVACIDRNGLQSCPIKEDSFLQRYSSDPTGALTED SIDDTFLPVPEYINQSVPKRPAGSVQNPVYHNQPLNPAPSRDPHYQDPHSTAVGNPEYLN TVQPTCVNSTFDSPAHWAQKGSHQISLDNPDYQQDFFPKEAKPNGIFKGSTAENAEYLRV APQSSEFIGA Thus, the present disclosure provides variants of human EGFR or a homologue thereof in which any one 15 or more of the amino acids in bold above are replaced with the indicated amino acids. EGFR variants of the disclosure further include variants in which a splice variant at L1038 or E1091 results in retention of an intronic sequence and premature termination of EGFR. The coding sequence of wild-type EGFR mRNA is reproduced below: >AY888105.1 Synthetic construct Homo sapiens clone FLH158490.01X 20 epidermal growth factor receptor (EGFR) mRNA, complete cds ATGCGACCCTCCGGGACGGCCGGGGCAGCGCTCCTGGCGCTGCTGGCTGCGCTCTGCCCGGCGAGTCGGGCT CTGGAGGAAAAGAAAGTTTGCCAAGGCACGAGTAACAAGCTCACGCAGTTGGGCACTTTTGAAGATCATTTT CTCAGCCTCCAGAGGATGTTCAATAACTGTGAGGTGGTCCTTGGGAATTTGGAAATTACCTATGTGCAGAGG AATTATGATCTTTCCTTCTTAAAGACCATCCAGGAGGTGGCTGGTTATGTCCTCATTGCCCTCAACACAGTG 25 GAGCGAATTCCTTTGGAAAACCTGCAGATCATCAGAGGAAATATGTACTACGAAAATTCCTATGCCTTAGCA GTCTTATCTAACTATGATGCAAATAAAACCGGACTGAAGGAGCTGCCCATGAGAAATTTACAGGAAATCCTG CATGGCGCCGTGCGGTTCAGCAACAACCCTGCCCTGTGCAACGTGGAGAGCATCCAGTGGCGGGACATAGTC AGCAGTGACTTTCTCAGCAACATGTCGATGGACTTCCAGAACCACCTGGGCAGCTGCCAAAAGTGTGATCCA AGCTGTCCCAATGGGAGCTGCTGGGGTGCAGGAGAGGAGAACTGCCAGAAACTGACCAAAATCATCTGTGCC 30 CAGCAGTGCTCCGGGCGCTGCCGTGGCAAGTCCCCCAGTGACTGCTGCCACAACCAGTGTGCTGCAGGCTGC ACAGGCCCCCGGGAGAGCGACTGCCTGGTCTGCCGCAAATTCCGAGACGAAGCCACGTGCAAGGACACCTGC CCCCCACTCATGCTCTACAACCCCACCACGTACCAGATGGATGTGAACCCCGAGGGCAAATACAGCTTTGGT GCCACCTGCGTGAAGAAGTGTCCCCGTAATTATGTGGTGACAGATCACGGCTCGTGCGTCCGAGCCTGTGGG GCCGACAGCTATGAGATGGAGGAAGACGGCGTCCGCAAGTGTAAGAAGTGCGAAGGGCCTTGCCGCAAAGTG 35 TGTAACGGAATAGGTATTGGTGAATTTAAAGACTCACTCTCCATAAATGCTACGAATATTAAACACTTCAAA AACTGCACCTCCATCAGTGGCGATCTCCACATCCTGCCGGTGGCATTTAGGGGTGACTCCTTCACACATACT CCTCCTCTGGATCCACAGGAACTGGATATTCTGAAAACCGTAAAGGAAATCACAGGGTTTTTGCTGATTCAG GCTTGGCCTGAAAACAGGACGGACCTCCATGCCTTTGAGAACCTAGAAATCATACGCGGCAGGACCAAGCAA CATGGTCAGTTTTCTCTTGCAGTCGTCAGCCTGAACATAACATCCTTGGGATTACGCTCCCTCAAGGAGATA 40 AGTGATGGAGATGTGATAATTTCAGGAAACAAAAATTTGTGCTATGCAAATACAATAAACTGGAAAAAACTG TTTGGGACCTCCGGTCAGAAAACCAAAATTATAAGCAACAGAGGTGAAAACAGCTGCAAGGCCACAGGCCAG GTCTGCCATGCCTTGTGCTCCCCCGAGGGCTGCTGGGGCCCGGAGCCCAGGGACTGCGTCTCTTGCCGGAAT GTCAGCCGAGGCAGGGAATGCGTGGACAAGTGCAACCTTCTGGAGGGTGAGCCAAGGGAGTTTGTGGAGAAC TCTGAGTGCATACAGTGCCACCCAGAGTGCCTGCCTCAGGCCATGAACATCACCTGCACAGGACGGGGACCA 45 GACAACTGTATCCAGTGTGCCCACTACATTGACGGCCCCCACTGCGTCAAGACCTGCCCGGCAGGAGTCATG GGAGAAAACAACACCCTGGTCTGGAAGTACGCAGACGCCGGCCATGTGTGCCACCTGTGCCATCCAAACTGC ACCTACGGATGCACTGGGCCAGGTCTTGAAGGCTGTCCAACGAATGGGCCTAAGATCCCGTCCATCGCCACT GGGATGGTGGGGGCCCTCCTCTTGCTGCTGGTGGTGGCCCTGGGGATCGGCCTCTTCATGCGAAGGCGCCAC ATCGTTCGGAAGCGCACGCTGCGGAGGCTGCTGCAGGAGAGGGAGCTTGTGGAGCCTCTTACACCCAGTGGA 50 GAAGCTCCCAACCAAGCTCTCTTGAGGATCTTGAAGGAAACTGAATTCAAAAAGATCAAAGTGCTGGGCTCC GGTGCGTTCGGCACGGTGTATAAGGGACTCTGGATCCCAGAAGGTGAGAAAGTTAAAATTCCCGTCGCTATC AAGGAATTAAGAGAAGCAACATCTCCGAAAGCCAACAAGGAAATCCTCGATGAAGCCTACGTGATGGCCAGC GTGGACAACCCCCACGTGTGCCGCCTGCTGGGCATCTGCCTCACCTCCACCGTGCAGCTCATCACGCAGCTC ATGCCCTTCGGCTGCCTCCTGGACTATGTCCGGGAACACAAAGACAATATTGGCTCCCAGTACCTGCTCAAC 55 TGGTGTGTGCAGATCGCAAAGGGCATGAACTACTTGGAGGACCGTCGCTTGGTGCACCGCGACCTGGCAGCC AGGAACGTACTGGTGAAAACACCGCAGCATGTCAAGATCACAGATTTTGGGCTGGCCAAACTGCTGGGTGCG GAAGAGAAAGAATACCATGCAGAAGGAGGCAAAGTGCCTATCAAGTGGATGGCATTGGAATCAATTTTACAC 28 AGAATCTATACCCACCAGAGTGATGTCTGGAGCTACGGGGTGACCGTTTGGGAGTTGATGACCTTTGGATCC AAGCCATATGACGGAATCCCTGCCAGCGAGATCTCCTCCATCCTGGAGAAAGGAGAACGCCTCCCTCAGCCA CCCATATGTACCATCGATGTCTACATGATCATGGTCAAGTGCTGGATGATAGACGCAGATAGTCGCCCAAAG TTCCGTGAGTTGATCATCGAATTCTCCAAAATGGCCCGAGACCCCCAGCGCTACCTTGTCATTCAGGGGGAT 5 GAAAGAATGCATTTGCCAAGTCCTACAGACTCCAACTTCTACCGTGCCCTGATGGATGAAGAAGACATGGAC GACGTGGTGGATGCCGACGAGTACCTCATCCCACAGCAGGGCTTCTTCAGCAGCCCCTCCACGTCACGGACT CCCCTCCTGAGCTCTCTGAGTGCAACCAGCAACAATTCCACCGTGGCTTGCATTGATAGAAATGGGCTGCAA AGCTGTCCCATCAAGGAAGACAGCTTCTTGCAGCGATACAGCTCAGACCCCACAGGCGCCTTGACTGAGGAC AGCATAGACGACACCTTCCTCCCAGTGCCTGAATACATAAACCAGTCCGTTCCCAAAAGGCCCGCTGGCTCT 10 GTGCAGAATCCTGTCTATCACAATCAGCCTCTGAACCCCGCGCCCAGCAGAGACCCACACTACCAGGACCCC CACAGCACTGCAGTGGGCAACCCCGAGTATCTCAACACTGTCCAGCCCACCTGTGTCAACAGCACATTCGAC AGCCCTGCCCACTGGGCCCAGAAAGGCAGCCACCAAATTAGCCTGGACAACCCTGACTACCAGCAGGACTTC TTTCCCAAGGAAGCCAAGCCAAATGGCATCTTTAAGGGCTCCACAGCTGAAAATGCAGAATACCTAAGGGTC GCGCCACAAAGCAGTGAATTTATTGGAGCATAG 15 Also described herein are mutations of EGFR that are associated with splice variants. All chromosomal locations are stated below with reference to NCBI RefSeq Assembly GCF_000001405.50 (GRCh38.p14). In an embodiment, a G to A mutation at location 7:55202626 results in disruption of the 5’ splice donor site in EGFR intron 27. Shown below is the sequence of EGFR exon 27, followed by the first 100 bases of intron 27 (underlined), including the disrupted canonical splice donor site (in a first pair of square brackets) with 20 the G to A mutation at location 7:55202626 (bold), and finally the observed alternative splice site (in a second pair of square brackets). In some embodiments, other intronic GT sites may act as alternative splice donor sites. <- intron 26 cont. CTGCAAAGCTGTCCCATCAAGGAAGACAGCTTCTTGCAGCGATACAGCTCAGACCCCACAGGCGCCTTGAC 25 TGAGGACAGCATAGACGACACCTTCCTCCCAGTGCCTG[AT]GAGTGGCTTGTCTGGAAACAGTCCTGCTC CTCAACCTCCTCGACCCACTCAGCAGCAGCCAGTCTCCAGTGTCCAAGCCAG[GT]GCTCCCTCCA -> intron 27 cont. As a result of the mutation in the canonical splice site, an alternative splice donor site is used in the downstream intron. The alternatively spliced mRNA therefore retains part of the intronic sequence after 30 EGFR exon 27, causing a frame shift in exon 28 (italic), and a premature stop codon resulting a C-terminal truncated form of EGFR. The full alternatively spliced mRNA transcript is presented below, with mutated splice donor site [AT], retained intronic sequence (underlined), position of the variant splice donor site (*), and frame shifted exon 28 (italic) where translation is terminated by a resulting premature stop codon (TAA, bold): 35 ATGCGACCCTCCGGGACGGCCGGGGCAGCGCTCCTGGCGCTGCTGGCTGCGCTCTGCCCGGCGAGTCGGGCT CTGGAGGAAAAGAAAGTTTGCCAAGGCACGAGTAACAAGCTCACGCAGTTGGGCACTTTTGAAGATCATTTT CTCAGCCTCCAGAGGATGTTCAATAACTGTGAGGTGGTCCTTGGGAATTTGGAAATTACCTATGTGCAGAGG AATTATGATCTTTCCTTCTTAAAGACCATCCAGGAGGTGGCTGGTTATGTCCTCATTGCCCTCAACACAGTG GAGCGAATTCCTTTGGAAAACCTGCAGATCATCAGAGGAAATATGTACTACGAAAATTCCTATGCCTTAGCA 40 GTCTTATCTAACTATGATGCAAATAAAACCGGACTGAAGGAGCTGCCCATGAGAAATTTACAGGAAATCCTG CATGGCGCCGTGCGGTTCAGCAACAACCCTGCCCTGTGCAACGTGGAGAGCATCCAGTGGCGGGACATAGTC AGCAGTGACTTTCTCAGCAACATGTCGATGGACTTCCAGAACCACCTGGGCAGCTGCCAAAAGTGTGATCCA AGCTGTCCCAATGGGAGCTGCTGGGGTGCAGGAGAGGAGAACTGCCAGAAACTGACCAAAATCATCTGTGCC CAGCAGTGCTCCGGGCGCTGCCGTGGCAAGTCCCCCAGTGACTGCTGCCACAACCAGTGTGCTGCAGGCTGC 45 ACAGGCCCCCGGGAGAGCGACTGCCTGGTCTGCCGCAAATTCCGAGACGAAGCCACGTGCAAGGACACCTGC CCCCCACTCATGCTCTACAACCCCACCACGTACCAGATGGATGTGAACCCCGAGGGCAAATACAGCTTTGGT GCCACCTGCGTGAAGAAGTGTCCCCGTAATTATGTGGTGACAGATCACGGCTCGTGCGTCCGAGCCTGTGGG GCCGACAGCTATGAGATGGAGGAAGACGGCGTCCGCAAGTGTAAGAAGTGCGAAGGGCCTTGCCGCAAAGTG TGTAACGGAATAGGTATTGGTGAATTTAAAGACTCACTCTCCATAAATGCTACGAATATTAAACACTTCAAA 29 AACTGCACCTCCATCAGTGGCGATCTCCACATCCTGCCGGTGGCATTTAGGGGTGACTCCTTCACACATACT CCTCCTCTGGATCCACAGGAACTGGATATTCTGAAAACCGTAAAGGAAATCACAGGGTTTTTGCTGATTCAG GCTTGGCCTGAAAACAGGACGGACCTCCATGCCTTTGAGAACCTAGAAATCATACGCGGCAGGACCAAGCAA CATGGTCAGTTTTCTCTTGCAGTCGTCAGCCTGAACATAACATCCTTGGGATTACGCTCCCTCAAGGAGATA 5 AGTGATGGAGATGTGATAATTTCAGGAAACAAAAATTTGTGCTATGCAAATACAATAAACTGGAAAAAACTG TTTGGGACCTCCGGTCAGAAAACCAAAATTATAAGCAACAGAGGTGAAAACAGCTGCAAGGCCACAGGCCAG GTCTGCCATGCCTTGTGCTCCCCCGAGGGCTGCTGGGGCCCGGAGCCCAGGGACTGCGTCTCTTGCCGGAAT GTCAGCCGAGGCAGGGAATGCGTGGACAAGTGCAACCTTCTGGAGGGTGAGCCAAGGGAGTTTGTGGAGAAC TCTGAGTGCATACAGTGCCACCCAGAGTGCCTGCCTCAGGCCATGAACATCACCTGCACAGGACGGGGACCA 10 GACAACTGTATCCAGTGTGCCCACTACATTGACGGCCCCCACTGCGTCAAGACCTGCCCGGCAGGAGTCATG GGAGAAAACAACACCCTGGTCTGGAAGTACGCAGACGCCGGCCATGTGTGCCACCTGTGCCATCCAAACTGC ACCTACGGATGCACTGGGCCAGGTCTTGAAGGCTGTCCAACGAATGGGCCTAAGATCCCGTCCATCGCCACT GGGATGGTGGGGGCCCTCCTCTTGCTGCTGGTGGTGGCCCTGGGGATCGGCCTCTTCATGCGAAGGCGCCAC ATCGTTCGGAAGCGCACGCTGCGGAGGCTGCTGCAGGAGAGGGAGCTTGTGGAGCCTCTTACACCCAGTGGA 15 GAAGCTCCCAACCAAGCTCTCTTGAGGATCTTGAAGGAAACTGAATTCAAAAAGATCAAAGTGCTGGGCTCC GGTGCGTTCGGCACGGTGTATAAGGGACTCTGGATCCCAGAAGGTGAGAAAGTTAAAATTCCCGTCGCTATC AAGGAATTAAGAGAAGCAACATCTCCGAAAGCCAACAAGGAAATCCTCGATGAAGCCTACGTGATGGCCAGC GTGGACAACCCCCACGTGTGCCGCCTGCTGGGCATCTGCCTCACCTCCACCGTGCAGCTCATCACGCAGCTC ATGCCCTTCGGCTGCCTCCTGGACTATGTCCGGGAACACAAAGACAATATTGGCTCCCAGTACCTGCTCAAC 20 TGGTGTGTGCAGATCGCAAAGGGCATGAACTACTTGGAGGACCGTCGCTTGGTGCACCGCGACCTGGCAGCC AGGAACGTACTGGTGAAAACACCGCAGCATGTCAAGATCACAGATTTTGGGCTGGCCAAACTGCTGGGTGCG GAAGAGAAAGAATACCATGCAGAAGGAGGCAAAGTGCCTATCAAGTGGATGGCATTGGAATCAATTTTACAC AGAATCTATACCCACCAGAGTGATGTCTGGAGCTACGGGGTGACTGTTTGGGAGTTGATGACCTTTGGATCC AAGCCATATGACGGAATCCCTGCCAGCGAGATCTCCTCCATCCTGGAGAAAGGAGAACGCCTCCCTCAGCCA 25 CCCATATGTACCATCGATGTCTACATGATCATGGTCAAGTGCTGGATGATAGACGCAGATAGTCGCCCAAAG TTCCGTGAGTTGATCATCGAATTCTCCAAAATGGCCCGAGACCCCCAGCGCTACCTTGTCATTCAGGGGGAT GAAAGAATGCATTTGCCAAGTCCTACAGACTCCAACTTCTACCGTGCCCTGATGGATGAAGAAGACATGGAC GACGTGGTGGATGCCGACGAGTACCTCATCCCACAGCAGGGCTTCTTCAGCAGCCCCTCCACGTCACGGACT CCCCTCCTGAGCTCTCTGAGTGCAACCAGCAACAATTCCACCGTGGCTTGCATTGATAGAAATGGGCTGCAA 30 AGCTGTCCCATCAAGGAAGACAGCTTCTTGCAGCGATACAGCTCAGACCCCACAGGCGCCTTGACTGAGGAC AGCATAGACGACACCTTCCTCCCAGTGCCTG[AT]GAGTGGCTTGTCTGGAAACAGTCCTGCTCCTCAACCT CCTCGACCCACTCAGCAGCAGCCAGTCTCCAGTGTCCAAGCCAG*AATACATAAACCAGTCCGTTCCCAAAA GGCCCGCTGGCTCTGTGCAGAATCCTGTCTATCACAATCAGCCTCTGAACCCCGCGCCCAGCAGAGACCCAC ACTACCAGGACCCCCACAGC -> exon 28 cont. 35 The protein sequence resulting from the alternative splicing at E1091 is reproduced below, with the retained intronic / partial frameshifted exon sequence (alternative residues 1091 – 1120) in bold: MRPSGTAGAALLALLAALCPASRALEEKKVCQGTSNKLTQLGTFEDHFLSLQRMFNNCEVVLGNLEITYVQR NYDLSFLKTIQEVAGYVLIALNTVERIPLENLQIIRGNMYYENSYALAVLSNYDANKTGLKELPMRNLQEIL HGAVRFSNNPALCNVESIQWRDIVSSDFLSNMSMDFQNHLGSCQKCDPSCPNGSCWGAGEENCQKLTKIICA 40 QQCSGRCRGKSPSDCCHNQCAAGCTGPRESDCLVCRKFRDEATCKDTCPPLMLYNPTTYQMDVNPEGKYSFG ATCVKKCPRNYVVTDHGSCVRACGADSYEMEEDGVRKCKKCEGPCRKVCNGIGIGEFKDSLSINATNIKHFK NCTSISGDLHILPVAFRGDSFTHTPPLDPQELDILKTVKEITGFLLIQAWPENRTDLHAFENLEIIRGRTKQ HGQFSLAVVSLNITSLGLRSLKEISDGDVIISGNKNLCYANTINWKKLFGTSGQKTKIISNRGENSCKATGQ VCHALCSPEGCWGPEPRDCVSCRNVSRGRECVDKCNLLEGEPREFVENSECIQCHPECLPQAMNITCTGRGP 45 DNCIQCAHYIDGPHCVKTCPAGVMGENNTLVWKYADAGHVCHLCHPNCTYGCTGPGLEGCPTNGPKIPSIAT GMVGALLLLLVVALGIGLFMRRRHIVRKRTLRRLLQERELVEPLTPSGEAPNQALLRILKETEFKKIKVLGS GAFGTVYKGLWIPEGEKVKIPVAIKELREATSPKANKEILDEAYVMASVDNPHVCRLLGICLTSTVQLITQL MPFGCLLDYVREHKDNIGSQYLLNWCVQIAKGMNYLEDRRLVHRDLAARNVLVKTPQHVKITDFGLAKLLGA EEKEYHAEGGKVPIKWMALESILHRIYTHQSDVWSYGVTVWELMTFGSKPYDGIPASEISSILEKGERLPQP 50 PICTIDVYMIMVKCWMIDADSRPKFRELIIEFSKMARDPQRYLVIQGDERMHLPSPTDSNFYRALMDEEDMD DVVDADEYLIPQQGFFSSPSTSRTPLLSSLSATSNNSTVACIDRNGLQSCPIKEDSFLQRYSSDPTGALTED SIDDTFLPVPDEWLVWKQSCSSTSSTHSAAASLQCPSQNT In an embodiment, a T to C mutation at location 7:55202627 results in disruption of the 5’ splice donor site in EGFR intron 27. Shown below is the sequence of EGFR exon 27, followed by the first 100 bases of intron 55 27 (underlined), including the disrupted canonical splice donor site (in a first pair of square brackets) with 30 the T to C mutation at location 7:55202627 (bold), and finally the alternative splice site (in a second pair of square brackets): <- intron 26 cont. CTGCAAAGCTGTCCCATCAAGGAAGACAGCTTCTTGCAGCGATACAGCTCAGACCCCACAGGCGCCTTGAC 5 TGAGGACAGCATAGACGACACCTTCCTCCCAGTGCCTG[GC]GAGTGGCTTGTCTGGAAACAGTCCTGCTC CTCAACCTCCTCGACCCACTCAGCAGCAGCCAGTCTCCAGTGTCCAAGCCAG[GT]GCTCCCTCCA -> intron 27 cont. As a result of the mutation in the canonical splice site, the alternative splice donor site is used in the downstream intron. The variant spliced mRNA therefore retains part of the intronic sequence after EGFR 10 exon 27, causing in a frame shift in exon 28 (italic), and a premature stop codon resulting a C-terminal truncated form of EGFR. The full alternatively spliced mRNA transcript is presented below, with mutated splice donor site [GC], retained intronic sequence (underlined), position of the variant splice donor site (*), and the upstream region of frame shifted exon 28 (italic) where translation is terminated by a resulting premature stop codon (TAA, bold): 15 ATGCGACCCTCCGGGACGGCCGGGGCAGCGCTCCTGGCGCTGCTGGCTGCGCTCTGCCCGGCGAGTCGGGCT CTGGAGGAAAAGAAAGTTTGCCAAGGCACGAGTAACAAGCTCACGCAGTTGGGCACTTTTGAAGATCATTTT CTCAGCCTCCAGAGGATGTTCAATAACTGTGAGGTGGTCCTTGGGAATTTGGAAATTACCTATGTGCAGAGG AATTATGATCTTTCCTTCTTAAAGACCATCCAGGAGGTGGCTGGTTATGTCCTCATTGCCCTCAACACAGTG GAGCGAATTCCTTTGGAAAACCTGCAGATCATCAGAGGAAATATGTACTACGAAAATTCCTATGCCTTAGCA 20 GTCTTATCTAACTATGATGCAAATAAAACCGGACTGAAGGAGCTGCCCATGAGAAATTTACAGGAAATCCTG CATGGCGCCGTGCGGTTCAGCAACAACCCTGCCCTGTGCAACGTGGAGAGCATCCAGTGGCGGGACATAGTC AGCAGTGACTTTCTCAGCAACATGTCGATGGACTTCCAGAACCACCTGGGCAGCTGCCAAAAGTGTGATCCA AGCTGTCCCAATGGGAGCTGCTGGGGTGCAGGAGAGGAGAACTGCCAGAAACTGACCAAAATCATCTGTGCC CAGCAGTGCTCCGGGCGCTGCCGTGGCAAGTCCCCCAGTGACTGCTGCCACAACCAGTGTGCTGCAGGCTGC 25 ACAGGCCCCCGGGAGAGCGACTGCCTGGTCTGCCGCAAATTCCGAGACGAAGCCACGTGCAAGGACACCTGC CCCCCACTCATGCTCTACAACCCCACCACGTACCAGATGGATGTGAACCCCGAGGGCAAATACAGCTTTGGT GCCACCTGCGTGAAGAAGTGTCCCCGTAATTATGTGGTGACAGATCACGGCTCGTGCGTCCGAGCCTGTGGG GCCGACAGCTATGAGATGGAGGAAGACGGCGTCCGCAAGTGTAAGAAGTGCGAAGGGCCTTGCCGCAAAGTG TGTAACGGAATAGGTATTGGTGAATTTAAAGACTCACTCTCCATAAATGCTACGAATATTAAACACTTCAAA 30 AACTGCACCTCCATCAGTGGCGATCTCCACATCCTGCCGGTGGCATTTAGGGGTGACTCCTTCACACATACT CCTCCTCTGGATCCACAGGAACTGGATATTCTGAAAACCGTAAAGGAAATCACAGGGTTTTTGCTGATTCAG GCTTGGCCTGAAAACAGGACGGACCTCCATGCCTTTGAGAACCTAGAAATCATACGCGGCAGGACCAAGCAA CATGGTCAGTTTTCTCTTGCAGTCGTCAGCCTGAACATAACATCCTTGGGATTACGCTCCCTCAAGGAGATA AGTGATGGAGATGTGATAATTTCAGGAAACAAAAATTTGTGCTATGCAAATACAATAAACTGGAAAAAACTG 35 TTTGGGACCTCCGGTCAGAAAACCAAAATTATAAGCAACAGAGGTGAAAACAGCTGCAAGGCCACAGGCCAG GTCTGCCATGCCTTGTGCTCCCCCGAGGGCTGCTGGGGCCCGGAGCCCAGGGACTGCGTCTCTTGCCGGAAT GTCAGCCGAGGCAGGGAATGCGTGGACAAGTGCAACCTTCTGGAGGGTGAGCCAAGGGAGTTTGTGGAGAAC TCTGAGTGCATACAGTGCCACCCAGAGTGCCTGCCTCAGGCCATGAACATCACCTGCACAGGACGGGGACCA GACAACTGTATCCAGTGTGCCCACTACATTGACGGCCCCCACTGCGTCAAGACCTGCCCGGCAGGAGTCATG 40 GGAGAAAACAACACCCTGGTCTGGAAGTACGCAGACGCCGGCCATGTGTGCCACCTGTGCCATCCAAACTGC ACCTACGGATGCACTGGGCCAGGTCTTGAAGGCTGTCCAACGAATGGGCCTAAGATCCCGTCCATCGCCACT GGGATGGTGGGGGCCCTCCTCTTGCTGCTGGTGGTGGCCCTGGGGATCGGCCTCTTCATGCGAAGGCGCCAC ATCGTTCGGAAGCGCACGCTGCGGAGGCTGCTGCAGGAGAGGGAGCTTGTGGAGCCTCTTACACCCAGTGGA GAAGCTCCCAACCAAGCTCTCTTGAGGATCTTGAAGGAAACTGAATTCAAAAAGATCAAAGTGCTGGGCTCC 45 GGTGCGTTCGGCACGGTGTATAAGGGACTCTGGATCCCAGAAGGTGAGAAAGTTAAAATTCCCGTCGCTATC AAGGAATTAAGAGAAGCAACATCTCCGAAAGCCAACAAGGAAATCCTCGATGAAGCCTACGTGATGGCCAGC GTGGACAACCCCCACGTGTGCCGCCTGCTGGGCATCTGCCTCACCTCCACCGTGCAGCTCATCACGCAGCTC ATGCCCTTCGGCTGCCTCCTGGACTATGTCCGGGAACACAAAGACAATATTGGCTCCCAGTACCTGCTCAAC TGGTGTGTGCAGATCGCAAAGGGCATGAACTACTTGGAGGACCGTCGCTTGGTGCACCGCGACCTGGCAGCC 50 AGGAACGTACTGGTGAAAACACCGCAGCATGTCAAGATCACAGATTTTGGGCTGGCCAAACTGCTGGGTGCG GAAGAGAAAGAATACCATGCAGAAGGAGGCAAAGTGCCTATCAAGTGGATGGCATTGGAATCAATTTTACAC AGAATCTATACCCACCAGAGTGATGTCTGGAGCTACGGGGTGACTGTTTGGGAGTTGATGACCTTTGGATCC AAGCCATATGACGGAATCCCTGCCAGCGAGATCTCCTCCATCCTGGAGAAAGGAGAACGCCTCCCTCAGCCA 31 CCCATATGTACCATCGATGTCTACATGATCATGGTCAAGTGCTGGATGATAGACGCAGATAGTCGCCCAAAG TTCCGTGAGTTGATCATCGAATTCTCCAAAATGGCCCGAGACCCCCAGCGCTACCTTGTCATTCAGGGGGAT GAAAGAATGCATTTGCCAAGTCCTACAGACTCCAACTTCTACCGTGCCCTGATGGATGAAGAAGACATGGAC GACGTGGTGGATGCCGACGAGTACCTCATCCCACAGCAGGGCTTCTTCAGCAGCCCCTCCACGTCACGGACT 5 CCCCTCCTGAGCTCTCTGAGTGCAACCAGCAACAATTCCACCGTGGCTTGCATTGATAGAAATGGGCTGCAA AGCTGTCCCATCAAGGAAGACAGCTTCTTGCAGCGATACAGCTCAGACCCCACAGGCGCCTTGACTGAGGAC AGCATAGACGACACCTTCCTCCCAGTGCCTG[GC]GAGTGGCTTGTCTGGAAACAGTCCTGCTCCTCAACCT CCTCGACCCACTCAGCAGCAGCCAGTCTCCAGTGTCCAAGCCAG*AATACATAAACCAGTCCGTTCCCAAAA GGCCCGCTGGCTCTGTGCAGAATCCTGTCTATCACAATCAGCCTCTGAACCCCGCGCCCAGCAGAGACCCAC 10 ACTACCAGGACCCCCACAGC -> exon 28 cont. The protein sequence resulting from the alternative splicing at E1091 is reproduced below, with the retained intronic / partial frameshifted exon sequence (alternative residues 1091 – 1120) in bold: MRPSGTAGAALLALLAALCPASRALEEKKVCQGTSNKLTQLGTFEDHFLSLQRMFNNCEVVLGNLEITYVQR NYDLSFLKTIQEVAGYVLIALNTVERIPLENLQIIRGNMYYENSYALAVLSNYDANKTGLKELPMRNLQEIL 15 HGAVRFSNNPALCNVESIQWRDIVSSDFLSNMSMDFQNHLGSCQKCDPSCPNGSCWGAGEENCQKLTKIICA QQCSGRCRGKSPSDCCHNQCAAGCTGPRESDCLVCRKFRDEATCKDTCPPLMLYNPTTYQMDVNPEGKYSFG ATCVKKCPRNYVVTDHGSCVRACGADSYEMEEDGVRKCKKCEGPCRKVCNGIGIGEFKDSLSINATNIKHFK NCTSISGDLHILPVAFRGDSFTHTPPLDPQELDILKTVKEITGFLLIQAWPENRTDLHAFENLEIIRGRTKQ HGQFSLAVVSLNITSLGLRSLKEISDGDVIISGNKNLCYANTINWKKLFGTSGQKTKIISNRGENSCKATGQ 20 VCHALCSPEGCWGPEPRDCVSCRNVSRGRECVDKCNLLEGEPREFVENSECIQCHPECLPQAMNITCTGRGP DNCIQCAHYIDGPHCVKTCPAGVMGENNTLVWKYADAGHVCHLCHPNCTYGCTGPGLEGCPTNGPKIPSIAT GMVGALLLLLVVALGIGLFMRRRHIVRKRTLRRLLQERELVEPLTPSGEAPNQALLRILKETEFKKIKVLGS GAFGTVYKGLWIPEGEKVKIPVAIKELREATSPKANKEILDEAYVMASVDNPHVCRLLGICLTSTVQLITQL MPFGCLLDYVREHKDNIGSQYLLNWCVQIAKGMNYLEDRRLVHRDLAARNVLVKTPQHVKITDFGLAKLLGA 25 EEKEYHAEGGKVPIKWMALESILHRIYTHQSDVWSYGVTVWELMTFGSKPYDGIPASEISSILEKGERLPQP PICTIDVYMIMVKCWMIDADSRPKFRELIIEFSKMARDPQRYLVIQGDERMHLPSPTDSNFYRALMDEEDMD DVVDADEYLIPQQGFFSSPSTSRTPLLSSLSATSNNSTVACIDRNGLQSCPIKEDSFLQRYSSDPTGALTED SIDDTFLPVPGEWLVWKQSCSSTSSTHSAAASLQCPSQNT In an embodiment, a G to A mutation at location 7:55201356 results in disruption of the 5’ splice donor site 30 in EGFR intron 25. Shown below is the sequence of EGFR exon 25, followed by the first 100 bases of intron 25 (underlined), including the disrupted canonical splice donor site (in a first pair of square brackets) with the G to A mutation at location 7:55201356 (bold), and a non-limiting number of predicted / potential alternative splice sites (in subsequent pairs of square brackets). Disruption of the canonical splice donor site results in retention of a length of intronic sequence comprising a first stop codon TGA (bold): 35 <- intron 24 GGGGATGAAAGAATGCATTTGCCAAGTCCTACAGACTCCAACTTCTACCGTGCCCTGATGGATGAAGAAGA CATGGACGACGTGGTGGATGCCGACGAGTACCTCATCCCACAGCAGGGCTTCTTCAGCAGCCCCTCCACGT CACGGACTCCCCTCCTGAGCTCTCTG[AT]ATGAAATCTCT[GT]CTCTCTCTCTCTCTCAAGCT[GT][G T]CTACTCATTTGAACAAATTGAATTTTAGGGAAAATAACCATCTA[GT]GAAACTCACATGGAT 40 -> intron 25 cont. As a result of the mutation in the canonical splice site, an alternative splice donor site is used in the downstream intron. The alternatively spliced mRNA therefore retains part of the intronic sequence after EGFR exon 25, causing frame shift of exons 26-28, and retention of a premature stop codon in the intron 25 sequence, resulting a C-terminal truncated form of EGFR. An example of a full alternatively spliced 45 mRNA transcript is presented below, with mutated splice donor site [AT], retained intronic sequence (underlined), position of the first variant splice donor site (*), where translation is terminated by a resulting premature stop codon (TAA, bold), and frame shifted exon 26 (italics). The use of downstream alternative splice donor sites will result in retention of increasingly long stretches of intronic sequence. ATGCGACCCTCCGGGACGGCCGGGGCAGCGCTCCTGGCGCTGCTGGCTGCGCTCTGCCCGGCGAGTCGGGC 50 TCTGGAGGAAAAGAAAGTTTGCCAAGGCACGAGTAACAAGCTCACGCAGTTGGGCACTTTTGAAGATCATT TTCTCAGCCTCCAGAGGATGTTCAATAACTGTGAGGTGGTCCTTGGGAATTTGGAAATTACCTATGTGCAG 32 AGGAATTATGATCTTTCCTTCTTAAAGACCATCCAGGAGGTGGCTGGTTATGTCCTCATTGCCCTCAACAC AGTGGAGCGAATTCCTTTGGAAAACCTGCAGATCATCAGAGGAAATATGTACTACGAAAATTCCTATGCCT TAGCAGTCTTATCTAACTATGATGCAAATAAAACCGGACTGAAGGAGCTGCCCATGAGAAATTTACAGGAA ATCCTGCATGGCGCCGTGCGGTTCAGCAACAACCCTGCCCTGTGCAACGTGGAGAGCATCCAGTGGCGGGA 5 CATAGTCAGCAGTGACTTTCTCAGCAACATGTCGATGGACTTCCAGAACCACCTGGGCAGCTGCCAAAAGT GTGATCCAAGCTGTCCCAATGGGAGCTGCTGGGGTGCAGGAGAGGAGAACTGCCAGAAACTGACCAAAATC ATCTGTGCCCAGCAGTGCTCCGGGCGCTGCCGTGGCAAGTCCCCCAGTGACTGCTGCCACAACCAGTGTGC TGCAGGCTGCACAGGCCCCCGGGAGAGCGACTGCCTGGTCTGCCGCAAATTCCGAGACGAAGCCACGTGCA AGGACACCTGCCCCCCACTCATGCTCTACAACCCCACCACGTACCAGATGGATGTGAACCCCGAGGGCAAA 10 TACAGCTTTGGTGCCACCTGCGTGAAGAAGTGTCCCCGTAATTATGTGGTGACAGATCACGGCTCGTGCGT CCGAGCCTGTGGGGCCGACAGCTATGAGATGGAGGAAGACGGCGTCCGCAAGTGTAAGAAGTGCGAAGGGC CTTGCCGCAAAGTGTGTAACGGAATAGGTATTGGTGAATTTAAAGACTCACTCTCCATAAATGCTACGAAT ATTAAACACTTCAAAAACTGCACCTCCATCAGTGGCGATCTCCACATCCTGCCGGTGGCATTTAGGGGTGA CTCCTTCACACATACTCCTCCTCTGGATCCACAGGAACTGGATATTCTGAAAACCGTAAAGGAAATCACAG 15 GGTTTTTGCTGATTCAGGCTTGGCCTGAAAACAGGACGGACCTCCATGCCTTTGAGAACCTAGAAATCATA CGCGGCAGGACCAAGCAACATGGTCAGTTTTCTCTTGCAGTCGTCAGCCTGAACATAACATCCTTGGGATT ACGCTCCCTCAAGGAGATAAGTGATGGAGATGTGATAATTTCAGGAAACAAAAATTTGTGCTATGCAAATA CAATAAACTGGAAAAAACTGTTTGGGACCTCCGGTCAGAAAACCAAAATTATAAGCAACAGAGGTGAAAAC AGCTGCAAGGCCACAGGCCAGGTCTGCCATGCCTTGTGCTCCCCCGAGGGCTGCTGGGGCCCGGAGCCCAG 20 GGACTGCGTCTCTTGCCGGAATGTCAGCCGAGGCAGGGAATGCGTGGACAAGTGCAACCTTCTGGAGGGTG AGCCAAGGGAGTTTGTGGAGAACTCTGAGTGCATACAGTGCCACCCAGAGTGCCTGCCTCAGGCCATGAAC ATCACCTGCACAGGACGGGGACCAGACAACTGTATCCAGTGTGCCCACTACATTGACGGCCCCCACTGCGT CAAGACCTGCCCGGCAGGAGTCATGGGAGAAAACAACACCCTGGTCTGGAAGTACGCAGACGCCGGCCATG TGTGCCACCTGTGCCATCCAAACTGCACCTACGGATGCACTGGGCCAGGTCTTGAAGGCTGTCCAACGAAT 25 GGGCCTAAGATCCCGTCCATCGCCACTGGGATGGTGGGGGCCCTCCTCTTGCTGCTGGTGGTGGCCCTGGG GATCGGCCTCTTCATGCGAAGGCGCCACATCGTTCGGAAGCGCACGCTGCGGAGGCTGCTGCAGGAGAGGG AGCTTGTGGAGCCTCTTACACCCAGTGGAGAAGCTCCCAACCAAGCTCTCTTGAGGATCTTGAAGGAAACT GAATTCAAAAAGATCAAAGTGCTGGGCTCCGGTGCGTTCGGCACGGTGTATAAGGGACTCTGGATCCCAGA AGGTGAGAAAGTTAAAATTCCCGTCGCTATCAAGGAATTAAGAGAAGCAACATCTCCGAAAGCCAACAAGG 30 AAATCCTCGATGAAGCCTACGTGATGGCCAGCGTGGACAACCCCCACGTGTGCCGCCTGCTGGGCATCTGC CTCACCTCCACCGTGCAGCTCATCACGCAGCTCATGCCCTTCGGCTGCCTCCTGGACTATGTCCGGGAACA CAAAGACAATATTGGCTCCCAGTACCTGCTCAACTGGTGTGTGCAGATCGCAAAGGGCATGAACTACTTGG AGGACCGTCGCTTGGTGCACCGCGACCTGGCAGCCAGGAACGTACTGGTGAAAACACCGCAGCATGTCAAG ATCACAGATTTTGGGCTGGCCAAACTGCTGGGTGCGGAAGAGAAAGAATACCATGCAGAAGGAGGCAAAGT 35 GCCTATCAAGTGGATGGCATTGGAATCAATTTTACACAGAATCTATACCCACCAGAGTGATGTCTGGAGCT ACGGGGTGACTGTTTGGGAGTTGATGACCTTTGGATCCAAGCCATATGACGGAATCCCTGCCAGCGAGATC TCCTCCATCCTGGAGAAAGGAGAACGCCTCCCTCAGCCACCCATATGTACCATCGATGTCTACATGATCAT GGTCAAGTGCTGGATGATAGACGCAGATAGTCGCCCAAAGTTCCGTGAGTTGATCATCGAATTCTCCAAAA TGGCCCGAGACCCCCAGCGCTACCTTGTCATTCAGGGGGATGAAAGAATGCATTTGCCAAGTCCTACAGAC 40 TCCAACTTCTACCGTGCCCTGATGGATGAAGAAGACATGGACGACGTGGTGGATGCCGACGAGTACCTCAT CCCACAGCAGGGCTTCTTCAGCAGCCCCTCCACGTCACGGACTCCCCTCCTGAGCTCTCTGAGTGCAACCA GCAACAATTCCACCGTGGCTTGCATTGATAGAAATGGGCTGCAAAGCTGTCCCATCAAGGAAGACAGCTTC TTGCAGCGATACAGCTCAGACCCCACAGGCGCCTTGACTGAGGACAGCATAGACGACACCTTCCTCCCAGT GCCTGGGGGATGAAAGAATGCATTTGCCAAGTCCTACAGACTCCAACTTCTACCGTGCCCTGATGGATGAA 45 GAAGACATGGACGACGTGGTGGATGCCGACGAGTACCTCATCCCACAGCAGGGCTTCTTCAGCAGCCCCTC CACGTCACGGACTCCCCTCCTGAGCTCTCTGATATGAAATCTCT*AGTGCAACCAGCAACAATTCCACCGT GGCTTGCATTGATAGAAATGGG -> exon 27 cont. The protein sequence resulting from the example alternative splicing at L1038 is reproduced below, with 50 the retained intronic / partial frameshifted exon sequence (alternative residue I1039) in bold underlined: MRPSGTAGAALLALLAALCPASRALEEKKVCQGTSNKLTQLGTFEDHFLSLQRMFNNCEVVLGNLEITYVQR NYDLSFLKTIQEVAGYVLIALNTVERIPLENLQIIRGNMYYENSYALAVLSNYDANKTGLKELPMRNLQEIL HGAVRFSNNPALCNVESIQWRDIVSSDFLSNMSMDFQNHLGSCQKCDPSCPNGSCWGAGEENCQKLTKIICA QQCSGRCRGKSPSDCCHNQCAAGCTGPRESDCLVCRKFRDEATCKDTCPPLMLYNPTTYQMDVNPEGKYSFG 55 ATCVKKCPRNYVVTDHGSCVRACGADSYEMEEDGVRKCKKCEGPCRKVCNGIGIGEFKDSLSINATNIKHFK NCTSISGDLHILPVAFRGDSFTHTPPLDPQELDILKTVKEITGFLLIQAWPENRTDLHAFENLEIIRGRTKQ HGQFSLAVVSLNITSLGLRSLKEISDGDVIISGNKNLCYANTINWKKLFGTSGQKTKIISNRGENSCKATGQ VCHALCSPEGCWGPEPRDCVSCRNVSRGRECVDKCNLLEGEPREFVENSECIQCHPECLPQAMNITCTGRGP DNCIQCAHYIDGPHCVKTCPAGVMGENNTLVWKYADAGHVCHLCHPNCTYGCTGPGLEGCPTNGPKIPSIAT 60 GMVGALLLLLVVALGIGLFMRRRHIVRKRTLRRLLQERELVEPLTPSGEAPNQALLRILKETEFKKIKVLGS 33 GAFGTVYKGLWIPEGEKVKIPVAIKELREATSPKANKEILDEAYVMASVDNPHVCRLLGICLTSTVQLITQL MPFGCLLDYVREHKDNIGSQYLLNWCVQIAKGMNYLEDRRLVHRDLAARNVLVKTPQHVKITDFGLAKLLGA EEKEYHAEGGKVPIKWMALESILHRIYTHQSDVWSYGVTVWELMTFGSKPYDGIPASEISSILEKGERLPQP PICTIDVYMIMVKCWMIDADSRPKFRELIIEFSKMARDPQRYLVIQGDERMHLPSPTDSNFYRALMDEEDMD 5 DVVDADEYLIPQQGFFSSPSTSRTPLLSSLI The term “KRAS” refers to human GTPase KRAS (also known as KRAS2, RASK2) or a homologue thereof. The protein sequence of human KRAS is available under Uniprot accession number P01116 and is incorporated herein in its entirety. The genomic sequence of human KRAS is available under NCBI GeneID 3845, and is incorporated herein in its entirety. The present disclosure provides a plurality of variants in the 10 KRAS protein sequence that are associated with altered response to one or more anti-cancer drugs. The plurality of variants include KRAS proteins that have any one or more of the following amino acid substitutions: E62K, E63K, K117E, K117R, D119G. These amino acids are highlighted in bold in the sequence of human KRAS reproduced below: 15 MTEYKLVVVGAGGVGKSALTIQLIQNHFVDEYDPTIEDSYRKQVVIDGETCLLDILDTAG QEEYSAMRDQYMRTGEGFLCVFAINNTKSFEDIHHYREQIKRVKDSEDVPMVLVGNKCDL PSRTVDTKQAQDLARSYGIPFIETSAKTRQRVEDAFYTLVREIRQYRLKKISKEEKTPGC VKIKKCIIM 20 Thus, the present disclosure provides variants of human KRAS or a homologue thereof in which any one or more of the amino acids in bold above are replaced with the indicated amino acids. The term “PIK3CA” refers to human Phosphatidylinositol 4,5-bisphosphate 3-kinase catalytic subunit alpha isoform (also known as Phosphoinositide 3-kinase alpha) or a homologue thereof. The protein sequence of human PIK3CA is available under Uniprot accession number P42336 and is incorporated herein in its 25 entirety. The genomic sequence of human PIK3CA is available under NCBI GeneID 5290, and is incorporated herein in its entirety. The present disclosure provides a plurality of variants in the PIK3CA protein sequence that are associated with altered response to one or more anti-cancer drugs. The plurality of variants include PIK3CA proteins that have any one or more of the following amino acid substitutions: E545K, E542K, E547K, E970G, Q969R, T972A. These amino acids are highlighted in bold in the sequence 30 of human PIK3CA reproduced below: MPPRPSSGELWGIHLMPPRILVECLLPNGMIVTLECLREATLITIKHELFKEARKYPLHQ 35 LLQDESSYIFVSVTQEAEREEFFDETRRLCDLRLFQPFLKVIEPVGNREEKILNREIGFA IGMPVCEFDMVKDPEVQDFRRNILNVCKEAVDLRDLNSPHSRAMYVYPPNVESSPELPKH IYNKLDKGQIIVVIWVIVSPNNDKQKYTLKINHDCVPEQVIAEAIRKKTRSMLLSSEQLK LCVLEYQGKYILKVCGCDEYFLEKYPLSQYKYIRSCIMLGRMPNLMLMAKESLYSQLPMD CFTMPSYSRRISTATPYMNGETSTKSLWVINSALRIKILCATYVNVNIRDIDKIYVRTGI 40 YHGGEPLCDNVNTQRVPCSNPRWNEWLNYDIYIPDLPRAARLCLSICSVKGRKGAKEEHC PLAWGNINLFDYTDTLVSGKMALNLWPVPHGLEDLLNPIGVTGSNPNKETPCLELEFDWF SSVVKFPDMSVIEEHANWSVSREAGFSYSHAGLSNRLARDNELRENDKEQLKAISTRDPL SEITEQEKDFLWSHRHYCVTIPEILPKLLLSVKWNSRDEVAQMYCLVKDWPPIKPEQAME LLDCNYPDPMVRGFAVRCLEKYLTDDKLSQYLIQLVQVLKYEQYLDNLLVRFLLKKALTN 45 QRIGHFFFWHLKSEMHNKTVSQRFGLLLESYCRACGMYLKHLNRQVEAMEKLINLTDILK QEKKDETQKVQMKFLVEQMRRPDFMDALQGFLSPLNPAHQLGNLRLEECRIMSSAKRPLW LNWENPDIMSELLFQNNEIIFKNGDDLRQDMLTLQIIRIMENIWQNQGLDLRMLPYGCLS IGDCVGLIEVVRNSHTIMQIQCKGGLKGALQFNSHTLHQWLKDKNKGEIYDAAIDLFTRS 34 CAGYCVATFILGIGDRHNSNIMVKDDGQLFHIDFGHFLDHKKKKFGYKRERVPFVLTQDF LIVISKGAQECTKTREFERFQEMCYKAYLAIRQHANLFINLFSMMLGSGMPELQSFDDIA YIRKTLALDKTEQEALEYFMKQMNDAHHGGWTTKMDWIFHTIKQHALN Thus, the present disclosure provides variants of human PIK3CA or a homologue thereof in which any 5 one or more of the amino acids in bold above are replaced with the indicated amino acids. The term “AKT1” refers to human RAC-alpha serine / threonine-protein kinase (also known as PKB, RAC) or a homologue thereof. The protein sequence of human AKT1 is available under Uniprot accession number P31749 and is incorporated herein in its entirety. The genomic sequence of human AKT1 is available under NCBI GeneID 207, and is incorporated herein in its entirety. The present disclosure provides a plurality of 10 variants in the AKT1 protein sequence that are associated with altered response to one or more anti-cancer drugs. The plurality of variants include AKT1 proteins that have any one or more of the following amino acid substitutions: E17K, G16E. These amino acids are highlighted in bold in the sequence of human AKT1 reproduced below: 15 MSDVAIVKEGWLHKRGEYIKTWRPRYFLLKNDGTFIGYKERPQDVDQREAPLNNFSVAQC QLMKTERPRPNTFIIRCLQWTTVIERTFHVETPEEREEWTTAIQTVADGLKKQEEEEMDF RSGSPSDNSGAEEMEVSLAKPKHRVTMNEFEYLKLLGKGTFGKVILVKEKATGRYYAMKI LKKEVIVAKDEVAHTLTENRVLQNSRHPFLTALKYSFQTHDRLCFVMEYANGGELFFHLS 20 RERVFSEDRARFYGAEIVSALDYLHSEKNVVYRDLKLENLMLDKDGHIKITDFGLCKEGI KDGATMKTFCGTPEYLAPEVLEDNDYGRAVDWWGLGVVMYEMMCGRLPFYNQDHEKLFEL ILMEEIRFPRTLGPEAKSLLSGLLKKDPKQRLGGGSEDAKEIMQHRFFAGIVWQHVYEKK LSPPFKPQVTSETDTRYFDEEFTAQMITITPPDQDDSMECVDSERRPHFPQFSYSASGTA Thus, the present disclosure provides variants of human AKT1 or a homologue thereof in which any one or 25 more of the amino acids in bold above are replaced with the indicated amino acids. The term “MAP2K1” refers to human Dual specificity mitogen-activated protein kinase kinase 1 (also known as MEK1, PRKMK1). The protein sequence of human MAP2K1 is available under Uniprot accession number Q02750 and is incorporated herein in its entirety. The genomic sequence of human MAP2K1 is available under NCBI GeneID 5604, and is incorporated herein in its entirety. The present disclosure 30 provides a plurality of variants in the MAP2K1 protein sequence that are associated with altered response to one or more anti-cancer drugs. The plurality of variants include MAP2K1 proteins that have any one or more of the following amino acid substitutions: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A. These amino acids are highlighted in bold in the sequence of human MAP2K1 reproduced below: 35 >sp|Q02750|MP2K1_HUMAN Dual specificity mitogen-activated protein kinase kinase 1 OS=Homo sapiens OX=9606 GN=MAP2K1 PE=1 SV=2 MPKKKPTPIQLNPAPDGSAVNGTSSAETNLEALQKKLEELELDEQQRKRLEAFLTQKQKV GELKDDDFEKISELGAGNGGVVFKVSHKPSGLVMARKLIHLEIKPAIRNQIIRELQVLHE CNSPYIVGFYGAFYSDGEISICMEHMDGGSLDQVLKKAGRIPEQILGKVSIAVIKGLTYL 40 REKHKIMHRDVKPSNILVNSRGEIKLCDFGVSGQLIDSMANSFVGTRSYMSPERLQGTHY SVQSDIWSMGLSLVEMAVGRYPIPPPDAKELELMFGCQVEGDAAETPPRPRTPGRPLSSY GMDSRPPMAIFELLDYIVNEPPPKLPSGVFSLEFQDFVNKCLIKNPAERADLKQLMVHAF IKRSDAEEVDFAGWLCSTIGLNQPSTPTHAAGV 35 Thus, the present disclosure provides variants of human MAP2K1 or a homologue thereof in which any one or more of the amino acids in bold above are replaced with the indicated amino acids. The term “MAP2K2” refers to Dual specificity mitogen-activated protein kinase kinase 2 (also known as MEK2, MKK2, PRKMK2). The protein sequence of human MAP2K2 is available under Uniprot accession 5 number P36507 and is incorporated herein in its entirety. The genomic sequence of human MAP2K2 is available under NCBI GeneID 5605, and is incorporated herein in its entirety. The present disclosure provides a plurality of variants in the MAP2K2 protein sequence that are associated with altered response to one or more anti-cancer drugs. The plurality of variants include MAP2K2 proteins that have any one or more of the following amino acid substitutions: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P. These 10 amino acids are highlighted in bold in the sequence of human MAP2K2 reproduced below: >sp|P36507|MP2K2_HUMAN Dual specificity mitogen-activated protein kinase kinase 2 OS=Homo sapiens OX=9606 GN=MAP2K2 PE=1 SV=1 MLARRKPVLPALTINPTIAEGPSPTSEGASEANLVDLQKKLEELELDEQQKKRLEAFLTQ KAKVGELKDDDFERISELGAGNGGVVTKVQHRPSGLIMARKLIHLEIKPAIRNQIIRELQ 15 VLHECNSPYIVGFYGAFYSDGEISICMEHMDGGSLDQVLKEAKRIPEEILGKVSIAVLRG LAYLREKHQIMHRDVKPSNILVNSRGEIKLCDFGVSGQLIDSMANSFVGTRSYMAPERLQ GTHYSVQSDIWSMGLSLVELAVGRYPIPPPDAKELEAIFGRPVVDGEEGEPHSISPRPRP PGRPVSGHGMDSRPAMAIFELLDYIVNEPPPKLPNGVFTPDFQEFVNKCLIKNPAERADL KMLTNHTFIKRSEVEEVDFAGWLCKTLRLNQPGTPTRTAV 20 Thus, the present disclosure provides variants of human MAP2K2 or a homologue thereof in which any one or more of the amino acids in bold above are replaced with the indicated amino acids. The term “BRAF” refers to human Serine / threonine-protein kinase B-raf (also known as BRAF1, RAFB1) or a homologue thereof. The protein sequence of human BRAF is available under Uniprot accession number P15056 and is incorporated herein in its entirety. The genomic sequence of human BRAF is 25 available under NCBI GeneID 673, and is incorporated herein in its entirety. The present disclosure provides a plurality of variants in the BRAF protein sequence that are associated with altered response to one or more anti-cancer drugs. The plurality of variants include BRAF proteins that have any one or more of the following amino acid substitutions: V480A, K499E, K499R. These amino acids are highlighted in bold in the sequence of human BRAF reproduced below: 30 MAALSGGGGGGAEPGQALFNGDMEPEAGAGAGAAASSAADPAIPEEVWNIKQMIKLTQEH IEALLDKFGGEHNPPSIYLEAYEEYTSKLDALQQREQQLLESLGNGTDFSVSSSASMDTV TSSSSSSLSVLPSSLSVFQNPTDVARSNPKSPQKPIVRVFLPNKQRTVVPARCGVTVRDS 35 LKKALMMRGLIPECCAVYRIQDGEKKPIGWDTDISWLTGEELHVEVLENVPLTTHNFVRK TFFTLAFCDFCRKLLFQGFRCQTCGYKFHQRCSTEVPLMCVNYDQLDLLFVSKFFEHHPI PQEEASLAETALTSGSSPSAPASDSIGPQILTSPSPSKSIPIPQPFRPADEDHRNQFGQR DRSSSAPNVHINTIEPVNIDDLIRDQGFRGDGGSTTGLSATPPASLPGSLTNVKALQKSP GPQRERKSSSSSEDRNRMKTLGRRDSSDDWEIPDGQITVGQRIGSGSFGTVYKGKWHGDV 40 AVKMLNVTAPTPQQLQAFKNEVGVLRKTRHVNILLFMGYSTKPQLAIVTQWCEGSSLYHH LHIIETKFEMIKLIDIARQTAQGMDYLHAKSIIHRDLKSNNIFLHEDLTVKIGDFGLATV KSRWSGSHQFEQLSGSILWMAPEVIRMQDKNPYSFQSDVYAFGIVLYELMTGQLPYSNIN NRDQIIFMVGRGYLSPDLSKVRSNCPKAMKRLMAECLKKKRDERPLFPQILASIELLARS LPKIHRSASEPSLNRAGFQTEDFSLYACASPKTPIQAGGYGAFPVH 36 Thus, the present disclosure provides variants of human BRAF or a homologue thereof in which any one or more of the amino acids in bold above are replaced with the indicated amino acids. Therapeutics and therapeutic methods The present disclosure describes variants that are associated with response to specific drugs or classes of 5 drugs. The drugs or class of drugs may be mitogen pathway inhibitors. The term “mitogen pathway inhibitors” refers to compounds or compositions that act by direct or indirect inhibition of one or more proteins in the mitogen pathway, such as e.g. BRAF, KRAS, MAP2K1, MAP2K2, EGFR, AKT1, and PI3K. A mitogen pathway inhibitor may be a small molecule or a large molecule, such as e.g. a protein, peptide, nucleic acid or combinations thereof. 10 The term “trametinib” refers to the allosteric MAP2K1 / 2 inhibitor Trametinib, with formula C26H23FIN5O4 available under CHEMBL ID CHEMBL2103875 and trametinib dimethyl sulfoxide with formula C28H29FIN5O5S available as CHEMBL2105741. The term also encompasses any other MAP2K1 / 2 inhibitor unless context indicates otherwise, and in particular any allosteric MAP2K1 / 2 inhibitor and / or any MAP2K1 / 2 inhibitor that has the same binding site or a binding site that overlaps with that of trametinib. In 15 particular, variants described herein as activating mutations would be expected to cause resistance to other MAP2K1 / 2 inhibitors regardless of binding site. Thus, variants indicated herein as associated with response to trametinib are assumed to be associated with response to any MAP2K1 / 2 inhibitor unless context indicates otherwise. Thus, also described herein are variants associated with response to a class of drugs comprising MAP2K1 / 2 inhibitors. Any of these variants may be associated with response to a class of 20 drugs comprising allosteric MAP2K1 / 2 inhibitors. Any of these variants may be associated with response to a class of drugs comprising allosteric MAP2K1 / 2 inhibitors that have the same binding site or an overlapping binding site as trametinib. In embodiments, variants indicated herein as associated with response to trametinib may be in the context of BRAF V600E positive cancers. The term “Dabrafenib” refers to the BRAF inhibitor Dabrafenib, with formula C23H20F3N5O2S2 available 25 under CHEMBL ID CHEMBL2028663 and Dabrafenib mesylate with formula C24H24F3N5O5S3 available as CHEMBL2105729. The term also encompasses any BRAF inhibitor, and in particular any ATP- competitive kinase inhibitor of BRAF, any BRAF inhibitor indicated for treatment of BRAF V600E-positive cancers and / or any BRAF inhibitor that has the same binding site or a binding site that overlaps with that of Dabrefinib. Thus, variants indicated herein as associated with response to dabrefinib are assumed to be 30 associated with response to any BRAF inhibitor unless context indicates otherwise. In embodiments, a variant associated with resistance to dabrafenib, such as e.g. BRAF K499E / R mutation is also associated with resistance to MEK1 / 2 inhibitors. Thus, also described herein are variants associated with response to a class of drugs comprising BRAF inhibitors. Any of these variants may be associated with response to a class of drugs comprising ATP-competitive kinase inhibitors of BRAF. Any of these variants may be 35 associated with response to a class of drugs comprising ATP-competitive kinase BRAF inhibitors that have the same binding site or an overlapping binding site as dabrefinib. In embodiments, variants indicated herein as associated with response to dabrafenib may be in the context of BRAF V600E positive cancers. 37 The term “Cetuximab” refers to the EGFR inhibitor Cetuximab, a recombinant chimeric human / mouse IgG1 monoclonal antibody available under CHEMBL ID CHEMBL1201577 and Cetuximab sarotalocan an antibody drug conjugate available as CHEMBL4298088. The term also encompasses any EGFR inhibitor, and in particular any competitive inhibitor of EGFR (i.e. any ligand that competitively inhibits the binding of 5 EGF to EGFR), and / or any EGFR inhibitor that has the same binding site or a binding site that overlaps with that of Cetuximab. Thus, variants indicated herein as associated with response to Cetuximab are assumed to be associated with response to any EGFR inhibitor unless context indicates otherwise, such as e.g. when the variant modifies the binding site of cetuximab. Thus, also described herein are variants associated with response to a class of drugs comprising EGFR inhibitors. Any of these variants may be 10 associated with response to a class of drugs comprising competitive inhibitors of EGFR. Any of these variants may be associated with response to a class of drugs comprising competitive inhibitors that have the same binding site or an overlapping binding site as Cetuximab. In embodiments, variants indicated herein as associated with response to cetuximab may be in the context of BRAF V600E positive cancers. The term “gefitinib” refers to the EGFR tyrosine kinase inhibitor gefitinib, with formula C22H24ClFN4O3 15 available under CHEMBL ID CHEMBL939. The term also encompasses any EGFR inhibitor, and in particular any tyrosine kinase inhibitor of EGFR, and / or any EGFR inhibitor that has the same binding site or a binding site that overlaps with that of gefitinib. Thus, variants indicated herein as associated with response to gefitinib are assumed to be associated with response to any EGFR inhibitor unless context indicates otherwise, such as e.g. when the variant modifies the binding site of gefitinib. Thus, also described 20 herein are variants associated with response to a class of drugs comprising EGFR inhibitors. Any of these variants may be associated with response to a class of drugs comprising tyrosine kinase inhibitors of EGFR. Any of these variants may be associated with response to a class of drugs comprising tyrosine kinase inhibitors that have the same binding site or an overlapping binding site as gefitinib. In embodiments, variants indicated herein as associated with response to gefitinib may be in the context of EGFR-dependent 25 cancers. In embodiments, variants indicated herein as associated with response to gefitinib may be in the context of cancers that have a EGFR amplification and / or exon19 deletion and / or L858R mutation. The term “osimertinib” refers to the EGFR tyrosine kinase inhibitor osimertinib, with formula C28H33N7O2 available under CHEMBL ID CHEMBL3353410, and osimertinib mesylate with formula C29H37N7O5S available under CHEMBL ID CHEMBL3545063. The term also encompasses any EGFR inhibitor, and in 30 particular any tyrosine kinase inhibitor of EGFR, and / or any EGFR inhibitor that has the same binding site or a binding site that overlaps with that of osimertinib. Thus, variants indicated herein as associated with response to osimertinib are assumed to be associated with response to any EGFR inhibitor unless context indicates otherwise, such as e.g. when the variant modifies the binding site of osimertinib. Thus, also described herein are variants associated with response to a class of drugs comprising EGFR inhibitors. 35 Any of these variants may be associated with response to a class of drugs comprising tyrosine kinase inhibitors of EGFR. Any of these variants may be associated with response to a class of drugs comprising tyrosine kinase inhibitors that have the same binding site or an overlapping binding site as osimertinib. In 38 embodiments, variants indicated herein as associated with response to gefitinib may be in the context of EGFR-dependent cancers. In embodiments, variants indicated herein as associated with response to osimertinib may be in the context of cancers that have a EGFR amplification and / or exon19 deletion and / or L858R mutation. 5 The term “pictilisib” refers to the pan-PI3K inhibitor pictilisib, with formula C23H27N7O3S2 available under CHEMBL ID CHEMBL521851. The term also encompasses any PI3K inhibitor, including PIK3CA and pan- PI3K inhibitors, and in particular any ATP-competitive inhibitor of PI3K or PIK3CA, and / or any PI3K or PIK3CA inhibitor that has the same binding site or a binding site that overlaps with that of pictilisib. Thus, variants indicated herein as associated with response to pictilisib are assumed to be associated with 10 response to any PI3K or PIK3CA inhibitor unless context indicates otherwise. Thus, also described herein are variants associated with response to a class of drugs comprising PI3K inhibitors. Any of these variants may be associated with response to a class of drugs comprising ATP competitive inhibitors of PI3K. Any of these variants may be associated with response to a class of drugs comprising ATP-competitive inhibitors that have the same binding site or an overlapping binding site as pictilisib. In embodiments, variants 15 indicated herein as associated with response to pictilisib may be in the context of BRAF V600E positive cancers. In embodiments, variants associated with response to pictilisib that are AKT1 variants may be associated with response to any PI3K or PIK3CA inhibitor. The term “sotorasib” refers to the KRAS G12C inhibitor sotorasib with formula C30H30F2N6O3 available under CHEMBL ID CHEMBL4535757. The term also encompasses any KRAS inhibitor, including in 20 particular any KRAS inhibitor indicated for the treatment of KRAS G12C positive cancers, any KRAS that binds to and maintains KRAS in an inactive conformation, and / or any KRAS inhibitor that has the same binding site or a binding site that overlaps with that of sotorasib. Thus, variants indicated herein as associated with response to sotorasib are assumed to be associated with response to any KRAS inhibitor unless context indicates otherwise. Thus, also described herein are variants associated with response to a 25 class of drugs comprising KRAS inhibitors. Any of these variants may be associated with response to a class of drugs comprising inhibitors of KRAS that bind to and maintain KRAS in an inactive conformation. Any of these variants may be associated with response to a class of drugs comprising inhibitors of KRAS that bind to KRAS 12C and maintain KRAS in an inactive conformation. Any of these variants may be associated with response to a class of drugs comprising KRAS inhibitors that have the same binding site 30 or an overlapping binding site as sotorasib. In embodiments, variants indicated herein as associated with response to sotorasib may be in the context of KRAS G12C positive cancers. The term “adagrasib” refers to the KRAS G12C inhibitor adagrasib with formula C32H35ClFN7O2 available under CHEMBL ID CHEMBL4594350. The term also encompasses any KRAS inhibitor, including in particular any KRAS inhibitor indicated for the treatment of KRAS G12C positive cancers, any KRAS that 35 binds to and maintains KRAS in an inactive conformation, and / or any KRAS inhibitor that has the same binding site or a binding site that overlaps with that of adagrasib. Thus, variants indicated herein as associated with response to adagrasib are assumed to be associated with response to any KRAS inhibitor 39 unless context indicates otherwise. In embodiments, such as e.g. embodiments involving variants of KRAS E98G, Q99R, I100V or K101E variants, the term may encompass adagrasib but not sotorasib. Thus, also described herein are variants associated with response to a class of drugs comprising KRAS inhibitors. Any of these variants may be associated with response to a class of drugs comprising inhibitors of KRAS that 5 bind to and maintain KRAS in an inactive conformation. Any of these variants may be associated with response to a class of drugs comprising inhibitors of KRAS that bind to KRAS 12C and maintain KRAS in an inactive conformation. Any of these variants may be associated with response to a class of drugs comprising KRAS inhibitors that have the same binding site or an overlapping binding site as adagrasib. In embodiments, variants indicated herein as associated with response to adagrasib may be in the context of 10 KRAS G12C positive cancers. The term “AKT inhibitor” encompasses any inhibitor of AKT1, AKT2 and / or AKT3, also known as protein kinase B (PKB), or PKB alpha, including pan-AKT inhibitors. This includes, for example, MK-2206 with formula C25H21N5O available under CHEMBL ID ChEMBL1079175; capivasertib, sold under the name Truqap, with formula C21H25ClN6O2 available under CHEMBL ID ChEMBL2325741; ipatasertib with 15 formula C24H32ClN5O2 available under CAS number 1001264-89-6; afuresertib with formula C18H17Cl2FN4OS available under CAS number 1047644-62-1; uprosertib with formula C18H16Cl2F2N4O2 available under CAS number 1047634-65-0. Without wishing to be bound by theory, the present inventors believe that variants that are associated with response to a particular drug, where the variant is “canonical” (likely to prevent drug binding) are also 20 associated with response to any drug that has the same binding site as the particular drug. Furtehr, variants that are described herein as “activating / “driver” / “drug addiction” variants may have some dependency on genetic context, but are expected to behave similarly in activating that pathway, or interacting with another oncogenic pathway. Thus, the exact genetic background in which the variant has been identified may not need to be present even in such cases, and genetic backgrunds with other alterations in the same pathway 25 may be associated with the same response. In embodiments, a subject who has been identified has having a variant associated with resistance to one or more drugs as described herein may be selected for treatment or treated with a genome editing technology to remove the variant. For example, a subject may be trained with a prime editing guide RNA (pegRNA) or a composition comprising such a guide RNA. Thus, also described herein is a prime editing 30 guide RNA comprising a guide sequence that specifically hybridises to a genomic sequence encoding a mitogen pathway protein variant as described herein. Also described herein is a prime editing guide RNA comprising a guide sequence that specifically hybridises to a genomic sequence encoding a mitogen pathway protein variant as described herein, for use in treating a subject who has been identified as having a mitogen pathway protein variant as described herein. Also described herein is a method of treating a 35 subject who has been identified as having a mitogen pathway protein variant as described herein, the method comprising administering to said subject a therapeutically effective amount of a prime editing guide RNA comprising a guide sequence that specifically hybridises to a genomic sequence encoding the mitogen 40 pathway protein variant. Also described herein is the use of a prime editing guide RNA comprising a guide sequence that specifically hybridises to a genomic sequence encoding a mitogen pathway protein variant as described herein, for use in the manufacture of a medicament. The medicament may be for treatment of cancer in a subject who has been identified as having the mitogen pathway protein variant. 5 Detection of a variant in a patient or a tumour of a patient may be performed using any method known in the art, including genomic, transcriptomic and proteomic detection methods applied to a sample obtained from the patient. A “sample” as used herein may be a cell or tissue sample, a biological fluid, an extract (e.g. a DNA extract obtained from the subject), from which genomic and / or transcriptomic and / or proteomic material can be 10 obtained for genomic and / or transcriptomic and / or proteomic analysis, such as genomic sequencing (e.g. whole genome sequencing, whole exome sequencing, targeted / panel sequencing), RNA sequencing (also referred to as “RNAseq” or “RNA-seq”), molecular counting, PCR, RT-PCR, microarrays, protein detection assays (e.g. using mass spectrometry, affinity based tests, protein arrays etc.). The sample may be a cell, tissue or biological fluid sample obtained from a subject (e.g. a biopsy). Such samples may be referred to 15 as “subject samples”. In particular, the sample may be a blood sample, or a tumour sample, or a sample derived therefrom. For the purposes of obtaining RNA sequence data, a “sample” as used herein may be a cell or tissue sample, or an extract (e.g. a RNA extract obtained from a subject) from which transcriptomic material can be obtained. For the purposes of obtaining DNA / genomic sequence data, a “sample” as used herein may be a cell or tissue sample, a biological fluid, an extract (e.g. a DNA extract obtained from the 20 subject), from which genomic material can be obtained for genomic analysis, such as genomic sequencing (e.g. whole genome sequencing, whole exome sequencing). The sample may be one which has been freshly obtained from a subject or may be one which has been processed and / or stored prior to genomic / transcriptomic / proteomic analysis (e.g. frozen, fixed or subjected to one or more purification, enrichment or extraction steps). The sample may be a cell or tissue culture sample. As such, a sample as 25 described herein may refer to any type of sample comprising cells or genomic and / or transcriptomic and / or proteomic material derived therefrom, whether from a biological sample obtained from a subject, or from a sample obtained from e.g. a cell line. Further, the sample may be transported and / or stored, and collection may take place at a location remote from the sequence data acquisition (e.g. sequencing) location, and / or any computer-implemented method steps described herein may take place at a location remote from the 30 sample collection location and / or remote from the sequence data acquisition (e.g. sequencing) location (e.g. the computer-implemented method steps may be performed by means of a networked computer, such as by means of a “cloud” provider).The samples used in methods of the present disclosure are typically samples comprising tumour cells (e.g. a tumour sample or sample comprising circulating tumour cells) or genetic material derived from tumour cells (such as e.g. cell free DNA or cell DNA and / or RNA extracted 35 from a sample comprising cells). Such as sample may be a “mixed sample”. A “mixed sample” refers to a sample that is assumed to comprise multiple cell types or genetic material derived from multiple cell types. Within the context of the present disclosure, a mixed sample is typically one that comprises tumour cells or 41 is assumed (expected) to comprise tumour cells, or genetic material derived from tumour cells, and normal cells or genetic material derived from normal cells. Genetic material can comprise genomic material (e.g. DNA) or transcriptomic material (e.g. RNA). Samples obtained from subjects, such as e.g. tumour samples, are typically mixed samples (unless they are subject to one or more purification and / or separation steps). 5 Typically, the sample comprises tumour cells and at least one non-tumour cell type (and / or genetic material derived therefrom). A “tumour sample” refers to a sample derived from or obtained from a tumour. Such samples may comprise tumour cells and normal (non-tumour) cells. The normal cells may comprise immune cells (such as e.g. lymphocytes), and / or other normal (non-tumour) cells (e.g. stromal cells). A tumour may be a solid tumour or a non-solid or haematological tumour. A tumour sample may be a primary tumour 10 sample, tumour-associated lymph node sample, or a sample from a metastatic site from the subject. A sample comprising tumour cells or genetic material derived from tumour cells may be a bodily fluid sample. Thus, the genetic material derived from tumour cells may be circulating tumour DNA or tumour DNA in exosomes. Instead or in addition to this, the sample may comprise circulating tumour cells. A mixed sample may be a sample of cells, tissue or bodily fluid that has been processed to extract genetic material (e.g. 15 DNA or RNA) and / or proteromic material. Methods for extracting genetic material from biological samples are known in the art. A mixed sample may have been subject to one or more processing steps that may modify the proportion of the multiple cell types or genetic material derived from the multiple cell types in the sample. For example, a mixed sample comprising tumour cells may have been processed to enrich the sample in tumour cells. Thus, a sample of purified tumour cells may be referred to as a “mixed sample” on 20 the basis that small amounts of other types of cells may be present, even if the sample may be assumed, for a particular purpose, to be pure (i.e. to have a tumour fraction of 1 or 100%). Thus, any method described herein may comprise one or more of: obtaining a sample from a subject, processing said sample to extract DNA, RNA and / or proteins, and detecting the presence of one or more variants in saod sample or extract using any method known in the art (such as e.g. sequencing, PCR, digital 25 PCR, RT-PCR, microarrays, etc.). A "cancer" can comprise any one or more of the following: acute lymphocytic leukemia (ALL), acute myeloid leukemia (AML), adrenocortical cancer, anal cancer, bladder cancer, blood cancer, bone cancer, brain tumor, breast cancer, cancer of the female genital system, cancer of the male genital system, central nervous system lymphoma, cervical cancer, childhood rhabdomyosarcoma, childhood sarcoma, chronic 30 lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), colon and rectal cancer, colon cancer, endometrial cancer, endometrial sarcoma, esophageal cancer, eye cancer, gallbladder cancer, gastric cancer, gastrointestinal tract cancer, hairy cell leukemia, head and neck cancer, hepatocellular cancer, Hodgkin's disease, hypopharyngeal cancer, Kaposi's sarcoma, kidney cancer, laryngeal cancer, leukemia, leukemia, liver cancer, lung cancer, malignant fibrous histiocytoma, malignant thymoma, melanoma, 35 mesothelioma, multiple myeloma, myeloma, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, nervous system cancer, neuroblastoma, non-Hodgkin's lymphoma, oral cavity cancer, oropharyngeal cancer, osteosarcoma, ovarian cancer, pancreatic cancer, parathyroid cancer, penile 42 cancer, pharyngeal cancer, pituitary tumor, plasma cell neoplasm, primary CNS lymphoma, prostate cancer, rectal cancer, respiratory system, retinoblastoma, salivary gland cancer, skin cancer, small intestine cancer, soft tissue sarcoma, stomach cancer, stomach cancer, testicular cancer, thyroid cancer, urinary system cancer, uterine sarcoma, vaginal cancer, vascular system, Waldenstrom's macroglobulinemia and 5 Wilms' tumor. In embodiments, the cancer is selected from: colorectal cancer (such as e.g. colorectal adenocarcinoma), Ewing’s sarcoma, and lung cancer (such as e.g. lung adenocarcinoma). In embodiments, the cancer is a cancer that has a KRAS G12C mutation, an EGFR amplification and / or exon19 deletion, a BRAF V600E mutation, or an EWS-FLI1 fusion. In embodiments, the cancer is a colorectal cancer with a BRAF V600E mutation. In embodiments, the cancer is a lung cancer with a KRAS 10 G12C mutation. In embodiments, the cancer is a lung cancer with EGFR amplification and / or exon19 deletion. In embodiments, the cancer is an Exing sarcoma with an EWS-FLI1 fusion. A cancer treatment may comprise administration of a chemotherapeutic agent. The active compound of a given chemotherapeutic agent may be provided in the form of a corresponding salt, solvate, or prodrug. In this specification reference to the chemotherapeutic agent includes reference to such forms.It may be 15 convenient or desirable to prepare, purify, and / or handle a corresponding salt of the active compound, for example, a pharmaceutically-acceptable salt. Examples of pharmaceutically acceptable salts are known in the art and discussed in e.g. Berge et al., 1977, "Pharmaceutically Acceptable Salts," J. Pharm. Sci., Vol. 66, pp. 1-19. Unless otherwise specified, a reference to a particular compound also include salt forms thereof. 20 The term “compound” or “active compound” refers to an active agent. An active agent may be a compound, set of compounds or particles that has an effect on a target molecule, cell or organism when it is put in contact with it. The active compound may be a small molecule (e.g. a chemotherapeutic agent), a large molecule (e.g. a nucleic acid therapeutic agent, peptide or protein, including but not limited to antigen binding molecules and affinity reagents), a cell or particule, a radiotherapeutic agent, and any combinations 25 thereof. It may be convenient or desirable to prepare, purify, and / or handle a corresponding solvate of the active compound. The term "solvate" is used herein in the conventional sense to refer to a complex of solute (e.g., active compound, salt of active compound) and solvent. If the solvent is water, the solvate may be conveniently referred to as a hydrate, for example, a mono-hydrate, a di-hydrate, a tri-hydrate, etc. 30 Unless otherwise specified, a reference to a particular compound also include solvate forms thereof. It may be convenient or desirable to prepare, purify, and / or handle the active compound in the form of a prodrug. The term "prodrug," as used herein, pertains to a compound which, when metabolised (e.g., in vivo), yields the desired active compound. Typically, the prodrug is inactive, or less active than the active compound, but may provide advantageous handling, administration, or metabolic properties. 35 Unless otherwise specified, a reference to a particular compound also include prodrugs thereof. 43 Medicaments and pharmaceutical compositions according to aspects of the present invention may be formulated for administration by a number of routes, including but not limited to, parenteral, intravenous, intra-arterial, intramuscular, intratumoural, oral and nasal. The medicaments and compositions may be formulated in fluid or solid form. Fluid formulations may be formulated for administration by injection to a 5 selected region of the human or animal body. Administration is preferably in a "therapeutically effective amount", this being sufficient to show benefit to the individual. The actual amount administered, and rate and time-course of administration, will depend on the nature and severity of the disease being treated. Prescription of treatment, e.g. decisions on dosage etc, is within the responsibility of general practitioners and other medical doctors, and typically takes 10 account of the disorder to be treated, the condition of the individual patient, the site of delivery, the method of administration and other factors known to practitioners. Examples of the techniques and protocols mentioned above can be found in Remington’s Pharmaceutical Sciences, 20th Edition, 2000, pub. Lippincott, Williams & Wilkins. A pharmaceutical composition as used herein is a composition at least one active compound, as defined 15 above, and optionally one or more other pharmaceutically acceptable ingredients well known to those skilled in the art, e.g., carriers, diluents, excipients, etc. If formulated as discrete units (e.g., tablets, etc.), each unit contains a predetermined amount (dosage) of the active compound. The term "pharmaceutically acceptable" as used herein pertains to compounds, ingredients, materials, compositions, dosage forms, etc., which are, within the scope of sound medical judgment, suitable for use 20 in contact with the tissues of the subject in question (e.g., human) without excessive toxicity, irritation, allergic response, or other problem or complication, commensurate with a reasonable benefit / risk ratio. Each carrier, diluent, excipient, etc. must also be "acceptable" in the sense of being compatible with the other ingredients of the formulation. A compound or composition may be administered alone or in combination with other treatments, either 25 simultaneously or sequentially. Methods according to the present invention may be performed, or products may be present, in vitro, ex vivo, or in vivo. The term “in vitro” is intended to encompass experiments with materials, biological substances, cells and / or tissues in laboratory conditions or in culture whereas the term “in vivo” is intended to encompass experiments and procedures with intact multi-cellular organisms. “Ex vivo” refers to 30 something present or taking place outside an organism, e.g. outside the human or animal body, which may be on tissue (e.g. whole organs) or cells taken from the organism. Where the method is performed in vitro it may comprise a high throughput screening assay. Test compounds used in the method may be obtained from a synthetic combinatorial peptide library, or may be synthetic peptides or peptide mimetic molecules. Other test compounds may comprise defined chemical 35 entities, oligonucleotides or nucleic acid ligands. 44 The term “patient” or “subject” are used herein interchangeably. A patient may refer to a subject that has been identified as having or being likely to have cancer. The patient may be any animal or human. The patient may be a non-human mammal (e.g. a model animal such as a mouse or rat, a pet such as a dog, cat or horse). The patient may be a human patient. The patient may be male or female. Unless indicated 5 otherwise, all references to proteins and variants refer to human proteins. References to proteins and variants encompass the human protein variant and a homologue thereof in a different organism, such as e.g. a non-human mammal. Methods of screening and drug design Also described herein is a method of screening for compounds that overcome resistance of a cell that 10 incorporates a mitogen pathway protein variant that is associated with response to one or more drugs, wherein the mitogen pathway protein variants are selected from: EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, 15 I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, K499E, K499R; PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A; and AKT1 variants selected from: E17K, G16E, wherein the presence of the one or more mitogen pathway protein variants are indicative of resistance to one or more of: MAP2K1 / 2 20 inhibitors, BRAF inhibitors, EGFR inhibitors, PI3K inhibitors, and KRAS inhibitors; and wherein the method comprises: contacting a cell that expresses said one or more mitogen pathway protein variants with one or more candidate compounds or compositions and comparing the effect of the one or more candidate compounds or compositions on said cells with one or more control conditions, and / or contacting said one or more mitogen pathway protein variants with one or more candidate compounds or compositions and 25 comparing the effect of the one or more candidate compounds or compositions on said mitogen pathway protein variant(s) with one or more control conditions, and determining the effect of said one or more candidate compounds or compositions on activity of the mitogen pathway protein variants. Also described herein is a method of screening for compounds that overcome resistance of a cell that incorporates a mitogen pathway protein variant that is associated with response to one or more drugs, 30 wherein the mitogen pathway protein variants are selected from: EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, 35 V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, K499E, K499R; PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A; and AKT1 variants selected from: E17K, G16E, wherein the presence of the one 45 or more mitogen pathway protein variants are indicative of resistance to one or more of: MAP2K1 / 2 inhibitors, BRAF inhibitors, EGFR inhibitors, PI3K inhibitors, and KRAS inhibitors; and wherein the method comprises: obtaining an in silico model of the mitogen pathway protein or a part thereof comprising the variant, and performing computational modelling to identify one or more candidate compounds that are 5 predicted to bind to and inhibit the mitogen pathway protein variant. In embodiments, the compounds or compositions used in any screening method described herein may comprise one or more mitogen pathway protein inhibitors and / or one or more components of a genome editing technology configured to target a genome comprising one or more mitogen pathway protein variants as described herein. The one or more mitogen pathway inhibitors may be small molecule inhibitors, or large 10 molecules (such as e.g. proteins or peptides, nucleic acids, etc). The one or more components of a genome editing technology may comprise a prime editing guide RNA (pegRNA). In some embodiments, the one or more candidate compounds comprise an antigen-binding molecule. For example,the antigen-binding molecule is capable of binding to a linear or conformational epitope of EGFR, MAP2K1 / 2, BRAF, PI3KCA, AKT1 or KRAS. In embodiments, the epitope comprises one or more of said mitogen pathway protein 15 variant positions. In some embodiments, determining the effect of said one or more candidate compounds or compositions comprises detecting a change in the proliferation rate of cells that expresses said one or more mitogen pathway protein variants in the presence of a candidate compound, whereby a reduction in proliferation rate of said cells compared to a control, or compared to the proliferation rate in the absence of said 20 candidate compound, indicates said candidate compound is an inhibitor of mutant mitogen pathway protein signalling, optionally further comprising selecting a subset of candidate compounds or compositions which cause a reduction in said proliferation rate. In some embodiments, detecting a change in the proliferation rate of cells can be assessed by nucleoside-analog incorporation (tritiated thymidine, 5-ethynyl-2′-deoxyuridine (EdU), 5-Bromo-2′-deoxyuridine (BrdU)), cell cycle-associated protein detection (Ki-67,25 phosphorylated-histone H3, proliferating cell nuclear antigen), use of cytoplasmic proliferation dyes (carboxyfluorescein diacetate succinimidyl ester (CFSE), cell trace violet), indirect measures of cell proliferation (cell imaging, ATP, MTT assay, Alamar blue assay, clonogenic assay), or any other method known in the art. These assays may monitor the number of cells over time, the number of cellular divisions, metabolic activity, or DNA synthesis. In some embodiments, the proliferation rate is measured with a 30 commercial proliferation assay or kit, such as incucyte (Sartorius) or CellTiter-Glo assay (Promega). In some embodiments, determining the effect of said one or more candidate compounds or compositions comprises determining the extent of phosphorylation of downstream mitogen pathway proteins, optionally via Western blotting, optionally further comprising selecting a subset of candidate compounds or compositions which cause a reduction in the extent of phosphorylation of downstream mitogen pathway 35 proteins. In some embodiments, whole cell protein is first isolated from a cell grown in the presence of one or more candidate compounds or compositions. In some embodiments, protein is transferred to a membrane before blotting with primary antibodies. In some embodiments, the membrane is blotted with 46 one or more primary antibodies targeting EGFR, p-EGFR, b-actin, p-ERK, and / or ERK total. In some embodiments, the EGFR total antibody targets the 1068 epitope (e.g. Cell Signaling Technology #2232) or 1020-1046 epitope (e.g. BD Biosciences #610017). In some embodiments, the secondary antibodies are conjugated to horseradish peroxidase or another reporter moiety known in the art. 5 In some embodiments, said mitogen pathway protein variant is EGFR, MAP2K1 / 2, BRAF, PI3KCA, or AKT1, determining the effect of said one or more candidate compounds or compositions comprises contacting said mitogen pathway protein variant with a candidate compound in the presence of a phosphorylation substrate and ATP and detecting a change in the amount of phosphorylation of said substrate, whereby a reduction of phosphorylation of said substrate compared to a control, or compared to 10 phosphorylation of the substrate in the absence of the candidate compound, indicates said candidate compound is an inhibitor of mutant mitogen pathway protein signaling, optionally further comprising selecting a subset of candidate compounds or compositions which cause a reduction phosphorylation of said substrate. In some embodiments, the phosphorylation substrate and / or ATP is, or is conjugated to, a reporting moiety. 15 In some embodiments, the reporting moiety is chromogenic, fluorescent, luminescent (including bioluminescent) or radioactive. In some embodiments, the ATP is a radioisotope, for example33P-ATP or 32P-ATP. In some embodiments, the ATP is a fluorescent analog of ATP, such as 2',3'-O-Trinitrophenyl- adenosine-5'-monophosphate (TNP-AMP), a MANT (Methyl-anthraniloyl) ATP, 2-aminopurine-riboside- triphosphate, or 2,6-diaminopurine-ribose-5’-triphosphate. 20 In some embodiments, the fluorescent reporting moiety is a BODIPY dye, fluorescein, FAM, FITC, TAMRA, Cy3, Cy5, Alexa FluorTM(Thermo Fisher), an Atto dye, or a fluorescent protein such as a green (GFP), enhanced green (EGFP), yellow (YFP, Citrine, Venus, YPet), blue (BFP, EBFP, Azurite, mKalama1) or cyan (ECFP, Cerulean, CyPet, mTurquoise2) fluorescent protein. In some embodiments, the reporting moiety is a chromogen, such as AMC, Dabcyl, Dnp, EDANS, FAM, TAMRA or an Atto dye. 25 In some embodiments, the phosphorylation substrate is a peptide. In some embodiments, the peptide is, or is a fragment of, a downstream kinase, such as an Src family kinase (SFK), protein tyrosine kinase 2 (Pyk2) or protein kinase A (PKA). In some embodiments, the downstream kinase is phospholipase C (PLC), Janus kinase (JAK), proto-oncogene c-Src, a son-of-sevenless protein (SOS) or other guanosine nucleotide exchange factor (GEF). In some embodiments, the peptide fragment preferably contains one or more 30 tyrosine residues. In some embodiments, the peptide is poly(4:1 Glu, Tyr), also known as polyE4Y1. In some embodiments, the peptide fragment is RRLIEDNEYTARG derived from v-Src (412-422) (Santa Cruz Biotechnology). In other embodiments, the reaction volume or cell is contacted with a secondary reporting moiety, for example a reporting moiety that directly or indirectly measures the turnover of ATP into ADP. In some 35 embodiments, the secondary reporting moiety is a luciferase enzyme, such as Ultra-GloTMLuciferase or Kinase-Glo® (Promega). 47 In some embodiments, the mitogen pathway protein variant is a KRAS variant, and the method further comprises contacting said said cell that expresses said one or more mitogen pathway protein variants, or said mitogen pathway protein variant, with a candidate compound in the presence of GTP and / or GDP and detecting a change in the amount of GTP and / or GDP, whereby a reduction in the detected change 5 compared to a control, or compared to the detected change in the absence of the candidate compound, indicates said candidate compound is an inhibitor of mutant mitogen pathway protein signaling, optionally further comprising selecting a subset of candidate compounds or compositions which cause a reduction in said detected change. KRAS is a GTPase frequently mutated in cancer and associated with poor disease prognosis. Mutated RAS 10 enzymes are locked in the activated GTP bound state which facilitates enhanced RAS signalling in cancer cells. This method can identify compounds that lock KRAS in an inactive “OFF” state, preventing GTP exchange. In some embodiments, GTP-bound KRAS is contacted with reporting GDP, such that the exchange of GTP for GDP results in a change of reporting by GDP. Alternatively, GDP-bound KRAS is contacted with 15 reporting GTP, such that the exchange of GDP for GTP results in a change of reporting by GTP. In some embodiments, the nucleotide exchange is facilitated by contacting the mixture with an SOS1 / SOS2 protein. In some embodiments, the GDP and / or GTP is, or is conjugated to, a reporting moiety. In some embodiments, the reporting moiety is chromogenic, fluorescent, luminescent (including bioluminescent) or radioactive. In some embodiments, the reporting moiety is a BODIPY dye, fluorescein, FAM, FITC, TAMRA, 20 Cy3, Cy5, Alexa FluorTM(Thermo Fisher), an Atto dye, or a fluorescent protein such as a green (GFP), enhanced green (EGFP), yellow (YFP, Citrine, Venus, YPet), blue (BFP, EBFP, Azurite, mKalama1) or cyan (ECFP, Cerulean, CyPet, mTurquoise2) fluorescent protein. In some embodiments, the reporting moiety is a chromogen, such as AMC, Dabcyl, Aminohexanoic acid (DNP), EDANS (5-((2- Aminoethyl)amino)naphthalene-1-sulfonic acid), FAM, TAMRA (5-carboxytetramethylrhodamine) or an Atto 25 dye. In other embodiments, the reaction volume or cell is contacted with a secondary reporting moiety, for example a reporting moiety that directly or indirectly measures the turnover of ATP into ADP. In some embodiments, the secondary reporting moiety is a luciferase enzyme, such as Ultra-GloTMLuciferase or Kinase-Glo® (Promega). 30 In embodiments, the methods further comprises formulating as a pharmaceutical compoisition a compound that has been identified as overcoming resistance of a cell that incorporates said mitogen pathway protein variant. Also described herein is a compound that overcomes resistance of a cell that incorporates a mitogen pathway protein variant that is associated with response to one or more drugs, wherein the mitogen 35 pathway protein variants are selected from: EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, 48 D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, 5 K499E, K499R; PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A AKT1 variants selected from: E17K, G16E. In embodiments, the compound is a candidate compound identified by one or more methods described herein, such as a method of screening for compounds that overcome resistance of a cell that incorporates a mitogen pathway protein variant that is associated with response to one or more drugs as described herein. 10 Also described herein is a pharmaceutical composition comprising a compound that overcomes resistance of a cell that incorporates a mitogen pathway protein variant that is associated with response to one or more drugs, wherein the mitogen pathway protein variants are selected from: EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, 15 K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, K499E, K499R; PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A AKT1 variants selected from: E17K, G16E. In embodiments, the compound 20 is a candidate compound identified by one or more methods described herein, such as a method of screening for compounds that overcome resistance of a cell that incorporates a mitogen pathway protein variant that is associated with response to one or more drugs as described herein. In embodiments, the pharmaceutical composition further comprises one or more pharmaceutical excipients. Also described herein is an antigen-binding molecule capable of binding to a linear or conformational 25 epitope of EGFR, MAP2K1 / 2, BRAF, PI3KCA, AKT1 or KRAS, wherein the epitope comprises one or more mitogen pathway protein variant selected from: EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, 30 L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, K499E, K499R; PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A AKT1 variants selected from: E17K, G16E. In embodiments, the antigen-binding molecule is a candidate compound identified by one or more methods described herein, such as a method of screening for 35 compounds that overcome resistance of a cell that incorporates a mitogen pathway protein variant that is associated with response to one or more drugs as described herein. 49 Also described herein is a prime editing guide RNA capable of binding to a genomic sequence comprising one or more mitogen pathway protein variants selected from: variants that are associated with a C-terminal truncation of EGFR, and in particular a truncation after amino acids E1091 or L1038; EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, 5 S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; BRAF variants selected from: V480A, K499E, K499R; PIK3CA variants selected from: 10 E545K, E542K, E547K, E970G, Q969R, T972A; and AKT1 variants selected from: E17K, G16E. *** The features disclosed in the foregoing description, or in the following claims, or in the accompanying drawings, expressed in their specific forms or in terms of a means for performing the disclosed function, or a method or process for obtaining the disclosed results, as appropriate, may, separately, or in any 15 combination of such features, be utilised for realising the invention in diverse forms thereof. While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the 20 spirit and scope of the invention. For the avoidance of any doubt, any theoretical explanations provided herein are provided for the purposes of improving the understanding of a reader. The inventors do not wish to be bound by any of these theoretical explanations. Any section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described. Throughout this specification, including the claims which follow, unless the context requires otherwise, the 25 word “comprise” and “include”, and variations such as “comprises”, “comprising”, and “including” will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps. It must be noted that, as used in the specification and the appended claims, the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. Ranges may be expressed herein as from “about” one particular 30 value, and / or to “about” another particular value. When such a range is expressed, another embodiment includes from the one particular value and / or to the other particular value. Similarly, when values are expressed as approximations, by the use of the antecedent “about,” it will be understood that the particular value forms another embodiment. The term “about” in relation to a numerical value is optional and means for example + / - 10%. 50 Examples Introduction The present work is based on the recognition that rapid, prospective and systematic functional annotation of variants would accelerate the discovery of drug resistance mechanisms and improve the effectiveness 5 of cancer treatments. Clustered, regularly interspaced protospacer repeat (CRISPR)-based gene editing approaches such as base editing can be used to directly interpret the function of variants of unknown significance (Cuella-Martin 2021, Hanna 2021, Kim 2022, Sanchez-Rivera 2022, Coelho 2023, Lue 2023a, Lue 2023b, Martin-Rufino 2023, Cooper 2024), and are useful tools to study genetic mechanisms of resistance to cytokines and small 10 molecule inhibitors (Coelho 2023, Sangree 2022, Bock 2022). Cytidine and adenine base editors employ a Cas9 nickase fused to a deaminase, facilitating the programmed installation of C>T and A>G SNVs in the genome at high efficiency in physiologically-relevant cell types (Lue 2023a, Lue 2023b, Martin-Rufino 2023, Cooper 2024, Anzalone 2020, Komor 2016, Gaudelli 2017, Xu 2021).Here, the inventors use base editing at scale to investigate genetic mechanisms of acquired resistance to molecularly-targeted cancer therapies, 15 identifying variants of unknown significance conferring drug resistance and drug sensitisation in cancer cells. EXAMPLE 1 – Base editing screens map functional domains in driving oncogenes. The inventors investigated drug resistance to 10 molecularly-targeted cancer drugs, which are currently approved by the Food and Drug Administration (FDA), or under clinical investigation (Fig.1a). The inventors 20 selected four cancer cell lines that are sensitive to these agents (Yang 2013) and harbour diverse oncogenic drivers (van der Meer 2019); H23 (lung; KRAS G12C; sensitive to adagrasib, sotorasib), PC9 (lung; EGFR amplification and exon19 deletion; sensitive to osimertinib, gefitinib), HT-29 (colon; BRAF V600E; sensitive to trametinib, dabrafenib-cetuximab, pictilisib), and MHH-ES-1 (Ewing sarcoma; EWS-FLI1 fusion; sensitive to olaparib, niraparib). The inventors mutagenised 11 cancer genes (AKT1, BCL2, BRAF, EGFR, KRAS, 25 MAP2K1, MAP2K2, MYC, PARP1, PARP2, PIK3CA) that are common drug targets or within the same signalling pathway as targeted oncogenes for interrogation with a guide RNA (gRNA) library (n = 22,816) tiling these genes and their 5’ and 3’ untranslated regions (UTRs) (Table 1). As controls, the inventors included non-targeting (n = 57), intergenic-targeting gRNAs (n = 168), and gRNAs predicted to introduce splice variants (Kluesner 2021) in non-essential (n = 87) and essential genes (Kim 30 2021, Pacini 2021) (n = 316). The gRNA library was introduced into cancer cell lines expressing doxycycline-inducible cytidine base editor (CBE) or adenine base editor (ABE) (Coelho 2023) with relaxed PAM requirements (Cas9-NGN) (Nishimasu 2018), to increase the saturation of targeted mutagenesis. The inventors analysed the potential functional effects of thousands of gene variants on drug resistance in parallel by performing base editing 35 screens with a proliferation read-out in the presence of targeted anti-cancer drugs from 46 independent pooled genetic screens (Fig.1a). Average z-scores for validated variants of interest are provided in Figure 51 9. Base editing screen replicates were highly correlated (Fig. 4), and control gRNAs targeting essential genes were significantly depleted, indicating efficient base editing (Fig. 1b, 17a). The inventors excluded gRNAs with a high off-target score from further analysis (1.2 % of gRNAs, Methods), which were mostly within repetitive regions of the KRAS 3’UTR and associated with a significant growth disadvantage in CBE 5 and ABE screens (Fig.17b). Cancer cell models had enhanced sensitivity to gRNAs targeting their specific driver oncogenes. PC9 cells were most sensitive to deleterious edits in EGFR, whereas HT-29 were most sensitive to deleterious edits in BRAF, and both were sensitive to targeting MYC, consistent with reported genetic dependencies and oncogene addiction (Behan 2019) (Fig.1c). Specifically, base editing mutagenesis across BRAF in HT-29 10 cells revealed depleted gRNAs predicted to install missense variants enriched in crucial functional domains, such as the RAS-binding domain and protein kinase domain (10.2 % and 59.2 % of deleterious BRAF missense variants with z-scores < -2, respectively) (Fig.1d). The inventors identified important sites of post- translational modification, such as the phosphorylation sites (Hornbeck 2015) S446, S365 and T753. Finally, the inventors identified a predicted gain of function missense variant at residue L505 (Fig. 1d), 15 which has been reported to co-occur with BRAF V600E, causes resistance to the BRAF inhibitor vemurafenib, and increases MAPK signalling (Wagenaar 2014). The data obtained highlight that base editing can provide insights into protein structure-function relationships including critical domains and residues, which could be valuable in drug discovery campaigns. EXAMPLE 2 – Variants modulating drug sensitivity cluster into four functional classes 20 The integration of 46 mutagenesis screens led to the identification of four functional classes of variants modulating drug sensitivity. The inventors classified these as; 1) variants that confer a proliferation advantage in the presence of drug but are deleterious in the absence of drug (“drug addiction variants”); 2) conferring a proliferation advantage only in the presence of drug (“canonical drug resistance variants”); 3) conferring a proliferation advantage in the presence and absence of drug (“driver variants”); and 4) 25 sensitising to drug (“drug-sensitising variants”). As an example, the inventors used base editing screens to investigate variants conferring resistance to the allosteric MEK1 / 2 inhibitor trametinib (Tian 2023, Flaherty 2012) in HT-29 cells. The inventors observed all four classes of variants modulating drug sensitivity (Fig. 2a); drug addiction variants (n = 10 gRNAs), canonical drug resistance variants (n = 30 gRNAs), driver variants (n = 24 gRNAs), and drug-sensitising 30 variants (n = 111 gRNAs). Of the 175 hit-scoring gRNAs from the trametinib screen, 0 were control gRNAs, implying a high signal-to-noise ratio. The inventors identified drug-sensitising variants for trametinib as causing loss of function in EGFR (Fig. 2a, z-score < -2, Methods), indicating that combination therapies targeting RAF-MEK and EGFR are effective in BRAF-mutant colorectal cancer (CRC), and consistent with the clinical approval EGFR and BRAF inhibitors in CRC (Prahallad 2012, Kopetz 2019, Huijberts 2020). 35 The inventors confirmed this genetic interaction in HT-29 cells in genome-wide CRISPR-Cas9 knock-out 52 screens in the presence of the BRAF inhibitor, dabrafenib, where knock-out of EGFR was the strongest sensitising hit (Fig.17c). The inventors subsequently investigated resistance to the combination of BRAF and EGFR inhibitors by performing a base editing screen in the presence of dabrafenib and cetuximab (Fig. 2b). Of the 81 hit- 5 scoring gRNAs from the dabrafenib plus cetuximab screens, 0 were control gRNAs. As expected, the inventors detected three main classes of drug resistance variants, with drug-sensitising variants now largely absent for the combination therapy (8 with the combination therapy vs 111 with trametinib). Collectively, these data demonstrate that base editing screens can reveal functionally distinct variants modulating drug sensitivity and highlight effective drug combinations. Further, although dabrafenib and cetuximab were not 10 tested individually, it is clear for at least some of these variants (including all “canonical” drug-binding mutations) that the variant would be associated with response to one of the two drugs alone (e.g. EGFR S464L would be associated to response to cetuxmab alone or on combination with dabrafenib). EXAMPLE 3 – Drug addiction and canonical drug resistance variants The inventors identified examples of drug addiction and canonical drug resistance variants for multiple 15 drugs. Canonical drug resistance variants such as MEK1 L115P (Delaney 2002) (Fig.2a) and EGFR S464L (Arena 2015) (Fig. 2b) were within the drug binding pockets for trametinib and cetuximab, respectively, consistent with direct disruption of drug binding (Fig. 2c). In contrast, drug addiction variants were predominantly activating mutations in oncogenes within the MAPK signalling pathway (e.g. KRAS Q61R / E62G, MEK2 Y134H) (Emery 2009). The deleterious effects of these variants in the absence of 20 inhibition with trametinib is consistent with overactivation of MAPK signalling leading to oncogene-induced senescence (Sale 2019 Zhu 1998). This phenotype is consistent with the mutual exclusivity of activating mutations within KRAS and the downstream kinase BRAF in patient CRC samples (p<0.001, Fig. 17d), highlighting that the drug addiction phenotype is dependent on the absence of pre-existing mutations in the cancer cell. Notably, MEK1 Q56R / K57E / R and F53L / S drug addiction variants have been previously 25 reported as conferring resistance to cetuximab in CRC patients (Siravegna 2015, Russo 2016) (Fig. 2a, 2b). The inventors validated the effect of BRAF / EGFR and MEK1 / 2 inhibitor drug resistance variants using arrayed proliferation assays and by analysing cell signalling. Overall, the inventors set out to validate 4 / 13 drug addiction and 3 / 36 canonical drug resistance gRNAs for dabrafenib and cetuximab, and 4 / 10 drug30 addiction and 2 / 30 canonical drug resistance gRNAs for trametinib. In line with the inventors’ high- throughput screening data, all of the analysed variants led to robust drug resistance (Fig. 2d). Canonical drug resistance variants did not have adverse effects on cell growth in the absence of drug, however, the drug addiction variants grew slower than controls in the absence of drug treatment. Western blotting analysis revealed that these variants elevated basal MAPK signalling as measured by p-ERK1 / 2, and were 35 associated with elevated p21 protein expression (Fig. 2e and Fig. 18a). Addition of dabrafenib and cetuximab reduced MAPK activity and p21 expression to near wild-type levels. HT-29 cells harbouring drug addiction variants showed an altered cell morphology and increased ß-galactosidase staining in the 53 absence of drug, indicating increased senescence, which was reversed with drug treatment (Fig. 2f and Fig. 18b). These data support the hypothesis that cancer cell clones harbouring drug addiction variants may be eliminated by implementing intermittent drug scheduling – so-called “drug holidays” (Sun 2014, Das Thakur 2013, Moriceau 2015). 5 EXAMPLE 4 – Driver variants conferring drug resistance Driver variants in orthogonal signalling pathways or in kinases downstream of the drug target can give rise to drug resistance in cancer. The inventors observed rare gain of function variants arising from gRNAs predicted to install missense variants in AKT1 and PIK3CA (Fig.3a). These sites resided in known mutation hot-spots in cancer; AKT1 E17K and PIK3CA E545K / E542K (Tate 2019). The inventors also detected rarer 10 driver variants at residues AKT1 D323 and PIK3CA C378 and E365 (Tate 2019), although known driver variants of residue H1047 of PIK3CA were not detected due to the editing and saturation constraints of the CBE and ABE NGN base editors. Variants conferring resistance to pictilisib, a pan-PI3K inhibitor under clinical development for solid tumours (Schoffski 2018), were rare in HT-29 base editing screens. Of 40 hit-scoring gRNAs from the pictilisib 15 screens, one was a control gRNA (2.5 %). Only driver variants in the downstream kinase, AKT1, conferred effective drug resistance (Fig. 3b). In contrast, driver variants in the drug target itself, PIK3CA, only conferred a proliferation advantage in the absence of drug, suggesting that PIK3CA activating variants remain sensitive to pictilisib. The inventors also compared the genetic mechanisms of acquired resistance to the recently FDA-approved 20 KRAS G12C inhibitors, sotorasib (Skoulidis 2021) and adagrasib (Janne 2022), in H23 KRASG12Clung cancer cells. Of the hit-scoring gRNAs from the sotorasib and adagrasib screens, 3 / 198 (1.51 %) and 1 / 224 (0.45 %) were control gRNAs, respectively. KRAS variants proximal to the drug binding pocket (e.g. R68G, D69G), gave cross-resistance to sotorasib and adagrasib, consistent with clinical findings (Awad 2021). Other known resistance variants, including alternative G12 alleles, were not detected in this analysis, 25 highlighting that base editing screens are not saturating and cannot install all variants. However, an advantage of the inventors’ approach is the ability to perform pathway-wide screens across entire endogenous genes. This allowed the inventors to detect activating variants in downstream kinases that confer resistance to sotorasib and adagrasib (Fig. 3c). Many of these activating variants in downstream kinases MEK1 (gene MAP2K1 - Y130H, F129S) and MEK2 (gene MAP2K2 - Y134H and Q60R, K61R / E) 30 conferred resistance to both inhibitors. The findings suggest that these gain of function mutations would also confer resistance to recently developed pan-KRAS inhibitors (Kim 2023). EXAMPLE 5 – Drug resistance variants to EGFR inhibitors The inventors investigated drug resistance to the EGFR tyrosine kinase inhibitors, gefitinib and osimertinib. Of the hit-scoring gRNAs from the gefitinib and osimertinib screens, 3 / 103 (2.91 %) and 0 / 81 were control 35 gRNAs, respectively. The variant conferring the strongest resistance to gefitinib was the archetypal EGFR T790M gate-keeper mutation, in line with clinical data (Pao 2005, Vaclova 2021) (Fig.5a). High sensitivity 54 to gefitinib and robust enrichment of the T790M variant led to uncharacteristically low correlation between three independent screen replicates for variants other than T790M (Fig. 4). The second-generation EGFR inhibitor, osimertinib, is designed to target EGFR T790M (Cross 2014). For osimertinib, drug resistance conferring gRNAs were predicted to introduce EGFR missense variants surrounding residues D1006 and 5 D1013-E1015, implicating these variants of unknown significance as drug resistance variants (Fig.5b). The activating mutation in PIK3CA, E542K also conferred resistance to osimertinib, in line with clinical findings (Chmielecki 2023a). However, the clinically-observed C797S resistance mutation (Chmielecki 2023a, Chmielecki 2023b, Thress 2015) was notably absent, as base editing was unable to introduce this specific T>A transversion mutation. Therefore, the inventors used prime editing (Anzalone 2020) as a 10 complementary gene editing approach to screen variant function in PC9 cells deficient in MLH1 (Fig.20a), which has been shown to increase prime editing efficiency (Chen 2021, Ferreira da Silva 2022). Firstly, the inventors confirmed the installation of EGFR C797S in PC9 MLH1 KO doxycycline-inducible prime editor cells in the presence of osimertinib (Fig. 20a). The inventors then used a focused primed editing gRNA (pegRNA) library of 162 pegRNAs (Mathis 2023) (three pegRNAs per variant) designed to 15 install all amino acid substitutions achievable with SNVs at six residues of interest from base editing screens of EGFR and including C797 (Table 2). pegRNA sequencing revealed a strong correlation between independent biological replicates, both with and without selection with osimertinib (Fig.20c).6 / 6 pegRNAs installing C797S were significantly enriched in the presence of osimertinib (Fig. 5c). Interestingly, EGFR C797 and T790 substitutions to chemically distinct residues were significantly depleted (Fig. 5c and Fig. 20 20c), including substitutions to lysine and arginine (charged), or proline and phenylalanine (steric effects), suggesting that particular variants of these key drug resistance residues disrupt EGFR function, and potentially explaining why they are never observed as drug resistance alleles. Although the inventors achieved the sensitivity required to detect loss-of-function effects, pegRNAs designed to install stop codons in EGFR were not depleted (0 / 3; Fig. 5c), indicating low prime editing efficiencies for some pegRNAs, and 25 potentially relating to the high copy number of EGFR in PC9 cells (van der Meer 2019). Overall this highlights the relative merits of prime editing and base editing technologies (Anzalone 2020). The inventors set out to validate drug resistance variants in the C-terminal regulatory region of EGFR in arrayed validation experiments using a stabilised, engineered prime editing gRNA (epegRNA) design (Nelson 2022) in co-competition assays with WT cells. Interestingly, EGFR D1012N, A1031D and A1013G 30 prime edited variants had a modest growth advantage over WT cells and C797C synonymous variant harbouring cells, but this was significantly enhanced in the presence of osimertinib (Fig.5d). This highlights the utility of base editing screens to capture sites functionally involved in drug sensitivity across entire proteins, distal to drug-binding and active sites, and for directing subsequent scanning at higher resolution using prime editing. 35 EXAMPLE 6 – EGFR C-terminal truncating variants sensitise to EGFR inhibitors Interestingly, several base editing gRNAs targeting EGFR significantly increased the sensitivity of PC9 cells to gefitinib and osimertinib, but had minimal effect on cell growth in the absence of drug treatment (Fig.5a 55 and 5b). The same variants conferred sensitivity to both drugs, and were predicted to install splice variants at residue E1091 of EGFR in the regulatory C-terminal domain. The inventors investigated the drug- sensitising variants in EGFR further by using arrayed cell proliferation assays. This approach validated the effects of gRNAs in PC9 CBE and ABE cell lines, as gRNAs introducing these splice variants significantly 5 increased cell death in response to EGFR inhibitors, gefitinib and osimertinib (Fig. 6a). Editing with these gRNAs also significantly increased sensitivity to chemically distinct EGFR inhibitors including erlotinib, lapatinib, and the antibody cetuximab (Fig. 21a). In contrast, the sensitivity of the edited cells to chemotherapy agents, such as cisplatin and paclitaxel, was largely unchanged, indicating an EGFR-specific mechanism of sensitisation (Fig.21a). Sequencing the edited locus revealed that editing by CBE and ABE 10 efficiently disrupted a splice donor site (Fig. 21b). PCR analysis of the cDNA from edited cells revealed retention of intronic sequence after EGFR exon 27 (Fig. 6b), caused by the use of an alternative splice donor in the downstream intron (Fig.21c). This led to a frame-shift and the introduction of a premature stop codon after residue E1091. Western blotting indicated similar expression of the WT and mutant EGFR proteins, but the migration of the mutant EGFR was consistent with a smaller protein lacking C-terminal 15 epitopes and phospho-epitopes (p-Y1148) (Fig. 6c). MAPK signalling in the EGFR mutant cell lines was not altered relative to wild-type controls, however, reduction of p-ERK levels by EGFR inhibitors was more profound in the mutant cells relative to the non-targeting gRNA controls (Fig.6c), consistent with increased sensitivity to EGFR inhibition. Analysis of the EGFR surface protein expression in WT and mutant EGFR PC9 cells revealed a significant increase in surface expression of the EGFR C-terminal truncated protein, 20 suggesting differences in inhibitor sensitivity could relate to EGFR localisation (Fig. 6d and Fig. 21d). This is consistent with the C-terminal cytoplasmic tail having a regulatory role for receptor internalisation (Chen 1989). Interestingly, non-small cell lung cancer patients harbouring rare tumour EGFR C-terminal truncations have respond to EGFR inhibition (Lovely 2021), warranting further investigation. Taken together, these data highlight a novel mechanism of drug sensitisation to EGFR inhibitors, which may inform 25 patient stratification and further our understanding of the mechanism of action of EGFR inhibitors. EXAMPLE 7 – Perturb-seq functionally defines drug resistant cell states To further understand cancer drug resistance mechanisms and the underlying signalling networks, the inventors set out to study the transcriptional programmes driven by drug resistant variants (Fig. 7a). The inventors recently described a modified version of perturb-seq (Dixit 2016, Replogle 2022), sc-SNV-seq 30 (Cooper 2024), to investigate transcriptional changes in single cells harbouring different endogenous variants installed with base editing. In this approach, iBAR barcodes (Zhu 2019) in gRNAs are used to cluster clonal groups of daughter cells with identical genotypes. Using this methodology, the inventors introduced a barcoded validation gRNA library (n = 451) into HT-29 CBE and ABE cells and treated them acutely with the combination of dabrafenib and cetuximab, which target BRAF and EGFR, respectively (Fig. 35 7a). This focused gRNA library performed as expected in proliferation screens in PC9 cells, displaying a significant correlation in gRNA effect size with larger base editing screens and independently validating drug resistance hits for gefitinib and osimertinib (Fig.22a). Furthermore, HT-29 cells harbouring essential- 56 targeting control gRNAs were depleted from single-cell RNA-seq data relative to non-targeting gRNAs, indicating efficient editing (Fig. 22b). In total, the inventors obtained 27,823 cells with confident gRNA assignments, with an average representation of 63 cells per gRNA for CBE, and 87 cells per gRNA for ABE, excluding essential-targeting controls (Fig.10-11). 5 The inventors compared the transcriptional response of groups of cells harbouring gRNAs that conferred drug resistance in proliferation screens to those with control and non-targeting gRNAs, revealing distinct transcriptional programmes (Fig.7b). Of the 35 gRNAs tested that conferred resistance to dabrafenib and cetuximab in base editing proliferation screens, 22 (62.86 %) elicited a significant transcriptional response vs non-targeting gRNA harbouring cells (Methods). Differential gene expression analysis of each targeting 10 gRNA relative to the non-targeting controls revealed four clusters of gRNAs, which segregated by variant class and the target gene (Fig. 22c), indicating that drug resistant variant classes drive transcriptionally distinct cell states. Differences in gene expression between cells containing different variant classes were more striking in the ABE dataset than the CBE dataset (Fig.22d), as the CBE dataset lacked drug addiction variants, which had the strongest transcriptional effect. Drug resistance genotypes had higher expression 15 of transcripts involved in cell cycle progression, such as CDC20, BUB3 and CDC6 (Fig. 22e), consistent with an increase in S phase, and a reduction in G1 occupancy in drug resistant cells (Fig. 22c; p-value < 1.5e-14 for CBE, p-value < 2.2e-16 for ABE, chi-squared test). Furthermore, pathway analyses of differentially expressed genes in drug resistant cells indicated significant enrichment in E2F target genes and genes controlling the G2M checkpoint (Fig.23a). 20 The inventors directly compared the transcriptional programmes driven by different drug resistance variant classes. Differentially expressed genes between drug resistance classes were strongly associated with EGFR and MAPK pathway signalling (Fig. 7d and 7e). Pathway analyses indicated that drug addiction variants elicited activation of similar signalling pathways to canonical drug resistance variants, but these were more extreme in drug addiction variants (Fig. 23b). Ranking variants based on their impact on gene 25 expression highlighted the greater impact of drug addiction variants (Fig. 23c and 23d), with only a few exceptional examples of canonical drug resistance variants that were closer to drug addiction in their transcriptional impact (e.g. KRAS E63K / E62K and KRAS K117R / E / D119G). These data highlight that perturb-seq can be used to assess the level of oncogenic signalling induced by different endogenous drug resistance variants. 30 EXAMPLE 8 – Drug resistance variants drive transcriptional signatures of immune evasion Interestingly, some drug resistance variants caused a significant reduction in the expression of genes pertaining to the JAK-STAT signalling pathway (Fig. 7d); this was particularly apparent in drug addiction variants (Fig.7e and Fig.23e). Transcripts critical for antigen presentation, such as B2M, were significantly downregulated by drug addiction variants in MAP2K1 (Fig.23e). B2M and HLA-A were also downregulated 35 by driver variants in PIK3CA (Fig. 7f). This is consistent with elevated RAS pathway signalling eliciting an immunosuppressive state (Tian 2023, Watterson 2023, Mugarza 2022, Coelho 2017), and implying that drug resistant cells can adopt immune-resistant signatures (Haas 2021). 57 In line with perturb-seq data, HT-29 cells with engineered drug resistance variants had a significant reduction in B2M and HLA protein expression, with the exception of the MAP2K2 Y134H drug addiction variant (1 / 5 variants; Fig. 24a). The inventors also engineered three drug addiction variants with base editing into a primary colorectal cancer organoid (Coelho 2023, Cattaneo 2020, Dijkstra 2018), CRC-9, 5 harbouring FBXW7 and TP53 driver mutations. Although the presence of drug addiction variants alone did not significantly reduce B2M and HLA expression in this model, the inventors observed a consistent increase in B2M and HLA protein expression following MAPK pathway inhibition with trametinib, in line with previous reports (El-Jawhari 2014, Ebert 2016, Frederick 2013, Mimura 2013, Sers 2009) (Fig. 24b). Induction of B2M and HLA expression was significantly attenuated in tumour organoids harbouring drug 10 addiction variants conferring drug resistance (Fig. 24b and 24c). Co-competition assays with GFP+WT control organoids showed a proliferation advantage in MAP2K1 S194P and MAP2K2 Y134H harbouring cancer cells, which was enhanced in the presence of MEK inhibition (Fig.24d). The inventors next tested whether modulation of MAPK signalling would directly affect tumour cell killing by patient-derived autologous, anti-tumour T cells. In co-culture experiments, the inventors observed robust and comparable 15 levels of T-cell mediated killing in all tumour organoid lines (Fig.24e). MEK inhibition significantly enhanced T cell-mediated cancer cell killing, except in tumour organoids harbouring the MAP2K1 S194P drug addiction variant (Fig.24e). To test whether the transcriptional cell states induced by drug resistance variants were predictive of patient treatment outcomes, the inventors compared the transcriptional profiles associated with drug resistance 20 variants with those identified in single-cell RNA-seq datasets from patient tumours (n = 23) from a recent phase-two clinical study of a BRAF, MEK and PD-1 inhibitor combination in patients with BRAF V600E CRC (Tian 2023). The inventors generated a progression-free survival (PFS) outcome score – a metric based on the Spearman rank correlation of differentially expressed genes in drug resistant cells from perturb-seq data, and the log-fold change of gene expression changes at 15 days post-treatment versus 25 pre-treatment in patients with PFS >6 months. Strikingly, the inventors observed a significantly lower PFS outcome score for drug addiction, driver and canonical drug resistant variants compared to control and non- targeting gRNAs (Fig. 7g). Drug addiction variants had the lowest PFS outcome score, consistent with signatures of higher MAPK pathway activation and reduced JAK-STAT signalling. In summary, the inventors highlight perturb-seq as a scalable approach to functionally classify cancer drug 30 resistance variants by their mechanism of action, and indicate specific variants whereby the mechanism of resistance to molecularly-targeted therapies may converge with resistance to anti-tumour immunity. EXAMPLE 9 – A drug resistance variant map indicates potential second-line therapies Base editing installs edits within a specific activity window, which is focused on nucleotides ~4-9 of the gRNA target sequence (Komor 2016, Gaudelli 2017, Pallaseni 2022). Therefore, the inventors set out to 35 determine the precise genomic variants installed for top-scoring drug resistance gRNAs from CBE and ABE base editing screens. In total, the inventors generated 46 isogenic HT-29 base edited cell lines, and performed next-generation sequencing of endogenous genomic loci before and after editing (Fig. 8a). 58 Amplicon sequencing demonstrated accurate and efficient editing exceeding 10 % variant allele frequency for 43 / 45 gRNAs, with a median editing efficiency of 91 % within the activity window (Table 8). The inventors observed significantly lower editing rates (43 %) outside of the activity window, and rare edits (6 / 139) that were not the expected C>T and A>G transition variants (Fig.25). These rare variants were consistent with 5 previously reported cytosine deamination by ABE within TCY contexts (Arbab 2020) (where Y denotes pyrimidine). A survey of COSMIC-curated drug response data (Tate 2019) and literature relating to clinical incidence of resistance to the drugs analysed in this study (Brammeld 2017, Awad 2021) revealed 88 amino acid positions in the screened target proteins associated with drug response data (Table 5A, Methods). 85 of 10 these had at least one gRNA predicted to target the amino acid position, 30 of which (35.29 %) had a concordant drug resistance phenotype in the present screening dataset specific to the reported drug and gene target. In addition, the inventors observed 252 edits at amino acid positions that had not been previously associated with altered drug response (e.g. MAP2K1 S194P, BRAF K499E / R; Tables 5B-5C). Of the genotyped drug resistance variants, only 39 % were in the drug targets themselves, highlighting the 15 potential of inhibiting orthogonal or downstream signalling pathways. Therefore, we explored whether verified drug resistance variants remained sensitive to other inhibitors tested in our screening dataset, representing possible alternative lines of therapy (Fig. 8b). Notably, the PIK3CA driver variants caused cross-resistance to multiple inhibitors including EGFR and MEK inhibitors, but remained sensitive to pictilisib. Furthermore, activating drug addiction variants conferred resistance to multiple inhibitors (e.g. 20 MEK2 Y134H; resistance to six inhibitors), but could be sensitive to drug holidays in the context of BRAF V600E (Fig. 2d). Conversely, canonical drug resistance variants only conferred resistance to a single inhibitor (e.g. MEK1 L115P; trametinib), reflecting the mechanism of action is strictly related to drug binding, and thus remains sensitive to unrelated MAPK pathway inhibitors, such as dabrafenib and cetuximab. KRAS inhibitors had broadly overlapping canonical drug resistance variants, with the exception of Q99R- 25 region alterations, which were associated with resistance to adagrasib, but not sotorasib, highlighting the subtly different binding modes of these inhibitors (Awad 2021). Interestingly, the inventors identified rare canonical drug resistance variants, such as BRAF K499E / R, which conferred resistance to four inhibitors, even though they did not significantly affect cell growth in the absence of drug treatment. Similarly, KRAS E63K / E62K conferred resistance to the combination of 30 dabrafenib and cetuximab, and KRAS K117R / E / D119G conferred resistance to gefitinib and dabrafenib plus cetuximab. Perturb-seq showed these variants had an intermediate transcriptional impact (Fig. 23d), consistent with a “Goldilocks” level of oncogenic signalling that is well tolerated, even in the absence of pathway inhibition. In summary, we categorise variants that confer resistance to multiple inhibitors, and highlight possible alternative inhibitors or treatment schedules that could be effective in treating drug- 35 resistant cancers (Fig.8b). 59 Discussion The inventors report a prospective genetic landscape of drug resistance mechanisms in cancer, one of the most comprehensive functional investigations of genetic drug resistance mechanisms to-date, comprising 10 drugs and profiling 11 cancer genes spanning common drug targets and oncogenic pathways. The 5 inventors identify known mechanisms of drug resistance (e.g. EGFR T790M – gefitinib (Pao 2005, Vaclova 2021); MEK1 K57 – cetuximab (Siravegna 2015, Russo 2016, Brammeld 2017)), as well as previously unreported variants of unknown significance causing drug resistance in vitro (Tables 5A-C, Fig. 8). Importantly, the inventors establish a framework for the functional classification of variants modulating drug sensitivity, which could inform clinical management (Fig.8b). 10 The data suggests variants modulating drug sensitivity can be divided into four phenotypic categories; drug addiction, canonical drug resistance, drivers, and drug-sensitising variants. Single-cell transcriptomics enabled deep functional classification of drug resistance variants based on their mechanisms of action and transcriptional impact. The reduction in JAK-STAT pathway activity and reduced expression of antigen presentation machinery induced by drug resistance variants implies an overlap between transcriptional 15 programmes driving drug resistance and immune evasion (Tian 2023, Watterson 2023, Mugarza 2022, Coelho 2017, Haas 2021). These data further support the rationale of combining effective targeted therapies with immune checkpoint blockade in cancer (Watterson 2023), and suggest immunotherapies could be more effective in treatment-naïve patients. Drug addiction variants displayed signatures of elevated MAPK pathway signalling in perturb-seq, and 20 slower growth in the absence of drug in screens and proliferation assays. Drug holidays have therefore been proposed to mitigate the emergence of these drug resistant cancer cell clones (Sun 2014, Das Thakur 2013, Moriceau 2015). However, the clinical use of intermittent dosing has been limited to-date, partly due to the lack of success of these trials in the absence of testing for the presence of such drug resistance variants (Algazi 2020). The database of drug addiction variants presented here, combined with the 25 increasingly routine use of longitudinal tracking of variants with circulating tumour DNA profiling, could enable the accurate stratification of patients for drug holidays (Sartore-Bianchi 2022). High-throughput endogenous gene editing accurately identified cancer driver variants, and mapped drug binding sites of multiple inhibitors with different modalities without prior structural information. Therefore, base editing screens could be useful to inform the design of new inhibitors targeting drug-resistant proteins 30 (e.g. EGFR T790M, EGFR S464L, EGFR C797S and MEK1 L115P), to identify promising combination treatments, and to verify drug mechanism of action. Furthermore, the inventors envisage that this approach will be important for understanding genetic mechanisms of acquired resistance to new molecules in the future, even before the emergence of resistance in the clinic, thus providing early insights to inform detection and combat drug resistance to improve treatment efficacy. 35 In particular, the present examples show for the first time an association between each of the variants in Table 5B and response to the corresponding drugs as indicated in Table 5B, i.e. MAP2K2 mutation at amino 60 acid position 134 resulting in protein change Y / H is associated with resistance to gefitinib, adagrasib, sotorasib, trametinib (drug addiction) and dabrafenib+cetuximab (drug addiction), etc. Note that unless indicated otherwise, response to any of the drugs tested is taken to encompass response to any drug to the same target as the drug tested, and also specifically any drug that has the same 5 mechanism of action and / or the same or an overlapping binding site as the drug tested. Thus, also described herein is the use of the presence of any of the mutations in Table 5B in the tumour of a subject, for predicting the response of the subject to a drug indicated in Table 5B as associated with resistance or sensitivity in the presence of the mutation, or a drug to the same target, to the same target with the same mode of action, or to the same target with the same or overlapping binding site. Specifically, 10 also described herein is the use of the presence of any of the mutations in Table 5B in the tumour of a subject associated in Table 5B with a drug resistance or drug sensitising variant class, for predicting the response of the subject to a drug indicated in Table 5B as associated with resistance or sensitivity in the presence of the mutation, or a drug to the same target, to the same target with the same mode of action, or to the same target with the same or overlapping binding site. In such embodiments, the predicted 15 response may be associated with the variant class indicated in Table 5B. Also described herein are methods comprising such uses and further comprising identifying a treatment for the subject based on the predicted response. Also described herein are methods of treating the subject with the identified treatment. Further, the present examples show for the first time, an association between each of the variants in Table 20 9 and response to the corresponding drugs as indicated in Table 9. Thus, also described herein is the use of the presence of any of the mutations in Table 9 in the tumour of a subject, for predicting the response of the subject to a drug indicated in Table 9 as associated with resistance or sensitivity in the presence of the mutation, or a drug to the same target, to the same target with the same mode of action, or to the same target with the same or overlapping binding site. 25 Further, the present examples also experimentally validate each of the following relationships: - an association between: (i) mutations of one or both of KRAS E62 and E63 to K and (ii) resistance to dabrafenib+cetuximab; - an association between (i) mutation of KRAS K117 to E or R and / or D119 to G and (ii) resistance to gefitinib; 30 - an association between (i) mutation of KRAS K117 to E, R and / or D119 to G and (ii) resistance to dabrafenib+cetuximab; - an association between: (i) mutation of MAP2K2 Y134 to H and (ii) resistance to dabrafenib+cetuximab, trametinib, sotorasib, adagrasib and gefitinib; - a drug addiction of tumour cells with a mutation of MAP2K2 Y134 to H, where the drug is 35 dabrafenib+cetuximab, trametinib, sotorasib, adagrasib and gefitinib; - an association between (i) mutation of MAP2K1 S194 to P and (ii) resistance to dabrafenib+cetuximab; - an association between (i) mutation of MAP2K1 S194 to P and (ii) resistance to trametinib; 61 - an association between (i) mutation of MAP2K1 S194 to P and (ii) resistance to sotorasib; - an association between (i) mutation of MAP2K1 S194 to P and (ii) resistance to adagrasib; - a drug addiction of tumour cells with a mutation of MAP2K1 S194 to P, where the drug is dabrafenib+cetuximab or trametinib; 5 - an association between (i) mutation of MAP2K1 L98 to P and / or I99 to Tand (ii) resistance to trametinib; - a drug addiction of tumour cells with a mutation of MAP2K1 L98 to P and / or I99 to T and, where the drug is trametinib; - an association between (i) mutations in EGFR causing a new splice variant to appear at residue 10 E1091 (premature stop codon after residue E1091) and (ii) sensitivity to EGFR inhibitors including gefitinib, osimertinib, erlotinib, lapatinib, and cetuximab; - an association between (i) mutation of BRAF K499 to E or R and (ii) resistance to trametinib and dabrafenib+cetuximab. As illustrated on Figure 8b, tumours with mutations of one or both of KRAS E62 and E63 to K may be 15 treated with a therapy that is not dabrafenib+cetuximab or gefitinib, such as e.g. trametinib. Further, tumours with mutation of KRAS K117 to E or R and / or D119 to G may be treated with a therapy that is not dabrafenib+cetuximab or gefitinib, such as e.g. trametinib. The present examples show for the first time, an association between (i) mutations of PIK3CA E545 to K, E542 to K, or E547 to K and (ii) resistance to each of dabrafenib+cetuximab, trametinib, osimertinib and 20 pictilisib. As illustrated on Figure 8b, tumours with mutations of PIK3CA E545 to K, E542 to K, or E547 to K may be treated with a therapy that is not dabrafenib+cetuximab, osimertinib or trametinib, such as e.g. pictilisib. The present examples show for the first time, an association between (i) mutations of PIK3CA E970 to G, Q969 to R and T972 to A and (ii) resistance to each of sotorasib, adagrasib, dabrafenib+cetuximab, and 25 trametinib. As illustrated on Figure 8b, tumours with mutations of PIK3CA E970 to G, Q969 to R and / or T972 to A may be treated with a therapy that is not sotorasib, adagrasib dabrafenib+cetuximab, or trametinib, such as e.g. pictilisib. The present examples show for the first time, an association between (i) mutations of AKT1 E17 to K and / or G16 to E and (ii) resistance to each of dabrafenib+cetuximab, trametinib and pictilisib. Tumours with 30 mutations of AKT1 E17 to K and / or G16 to E may be treated with therapy that is not dabrafenib+cetuximab, trametinib or pictilisib, such as e.g. an AKT inhibitor. The present examples show for the first time, an association between: (i) mutation of MAP2K2 Y134 to H and (ii) resistance to dabrafenib+cetuximab, trametinib, sotorasib, adagrasib and gefitinib. Data for at least dabrafenib+cetuximab and trametinib indicates that the variant likely causes a drug addiction to these drugs 35 . As illustrated on Figure 8b, tumours with a mutation of MAP2K2 Y134 to H may be treated with any one 62 or more of dabrafenib+cetuximab, trametinib, sotorasib, adagrasib and gefitinib using an intermittent treatment scheme (also referred to as “drug holiday”). The present examples show for the first time, an association between: (i) mutation of MAP2K2 V195 to A and / or S198 to P and (ii) resistance to each of sotorasib and adagrasib. As illustrated on Figure 8b, tumours 5 with mutations of MAP2K2 V195 to A and / or S198 to P may be treated with a therapy that is not sotorasib or adagrasib. The present examples show for the first time, an association between: (i) mutation of MAP2K2 I208 to T and / or L210 to P and (ii) resistance to each of sotorasib and adagrasib. As illustrated on Figure 8b, tumours with mutations of MAP2K2 I208 to T and / or L210 to P may be treated with a therapy that is not sotorasib 10 or adagrasib. The present examples show for the first time, an association between: (i) mutation of BRAF V480 to A and (ii) resistance to dabrafenib+cetuximab. As illustrated on Figure 8b, tumours with mutations of BRAF V480 to A may be treated with a therapy that is not dabrafenib+cetuximab. The present examples show for the first time, an association between: (i) mutation of BRAF K499 to E or 15 R and (ii) resistance to each of dabrafenib+cetuximab and trametinib. As illustrated on Figure 8b, tumours with mutations of BRAF K499 to E or R may be treated with a therapy that is not dabrafenib+cetuximab or trametinib. The present examples show for the first time, an association between: (i) mutation of EGFR D46 to N and (ii) resistance to each of dabrafenib+cetuximab and trametinib. As illustrated on Figure 8b, tumours with 20 mutations of EGFR D46 to N may be treated with a therapy that is not dabrafenib+cetuximab or trametinib. The present examples show for the first time, an association between: (i) mutation of EGFR F436 to S or L and / or S437 to P and / or FS to PP at positions 436-437 and (ii) resistance to dabrafenib+cetuximab. As illustrated on Figure 8b, tumours with mutations of EGFR F436 to S or L and / or S437 to P and / or FS to PP at positions 436-437 may be treated with a therapy that is not dabrafenib+cetuximab, such as e.g. 25 trametinib. The present examples show for the first time, an association between: (i) mutation of EGFR S720 to F (ii) resistance to osimertinib. As illustrated on Figure 8b, tumours with mutations of EGFR S720 to F may be treated with a therapy that is not osimertinib, such as e.g. gefitinib. The present examples show for the first time, an association between: (i) mutation of EGFR D1006 to N 30 and / or E1005 to K and / or E1004 to K (ii) resistance to osimertinib. As illustrated on Figure 8b, tumours with mutations of EGFR D1006 to N and / or E1005 to K and / or E1004 to K may be treated with a therapy that is not osimertinib, such as e.g. gefitinib. The present examples show for the first time, an association between: (i) mutation of EGFR D1012 to N, D1014 to N, and / or E1015 to K (ii) resistance to osimertinib. As illustrated on Figure 8b, tumours with 63 mutations of EGFR D1012 to N, D1014 to N, and / or E1015 to K may be treated with a therapy that is not osimertinib, such as, e.g. gefitinib. The present examples show for the first time, an association between: (i) mutation of EGFR L1038 leading to a new splice site resulting in truncation of the EGFR protein after position 1038 and (ii) resistance to 5 osimertinib. As illustrated on Figure 8b, tumours with mutations of EGFR L1038 leading to a new splice site may be treated with a therapy that is not osimertinib, such as e.g. gefitinib. The present examples show for the first time, an association between: (i) mutation of EGFR E1091 leading to a new splice site resulting in truncation of the EGFR protein after position 1091 and (ii) sensitivity to each of gefitinib and osimertinib. As illustrated on Figure 8b, tumours with mutations of EGFR 1091 leading to a 10 new splice site resulting in truncation of the EGFR protein after position 1091 may be treated with a therapy that includes gefitinib and / or osimertinib. The present examples show for the first time, an association between: (i) mutation of MAP2K1 Q45 to R and / or Q46 to R and (ii) resistance to each of trametinib, dabrafenib+cetuximab and adagrasib. As illustrated on Figure 8b, tumours with mutations of MAP2K1 Q45 to R and / or Q46 to R may be treated with 15 a therapy that is not trametinib, dabrafenib+cetuximab or adagrasib. The present examples show for the first time, an association between: (i) mutation of MAP2K1 F53 to L or S and (ii) resistance to adagrasib. As illustrated on Figure 8b, tumours with mutations of MAP2K1 F53 to L or S may be treated with a therapy that is not trametinib, dabrafenib+cetuximab or adagrasib, or may be treated with an intermittent schedule of one or more of trametinib, dabrafenib+cetuximab or adagrasib. 20 The present examples show for the first time, an association between: (i) mutation of MAP2K1 Q56 to R and / or K57 to E or R and / or Q58 to R and (ii) resistance to each of trametinib, dabrafenib+cetuximab, sotorasib and adagrasib. As illustrated on Figure 8b, tumours with mutations of MAP2K1 K57 to E or R and / or Q58 to R may be treated with a therapy that is not trametinib, dabrafenib+cetuximab, sotorasib or adagrasib, or may be treated with an intermittent schedule of one or more of trametinib, 25 dabrafenib+cetuximab, sotorasib or adagrasib. The present examples show for the first time, an association between: (i) mutation of MAP2K1 I99 to T and / or L98 to P and (ii) resistance to trametinib. The data further indicates that these variants likely causes a drug addiction to trametinib. As illustrated on Figure 8b, tumours with mutations of MAP2K1 I99 to T and / or L98 to P may be treated with a therapy that is not trametinib, or with an intermittent treatment 30 schedule of trametinib (drug holiday). The present examples show for the first time, an association between: (i) mutation of MAP2K1 I111 to T and / or Q110 to H and / or I112 to T and (ii) resistance to each of trametinib, and dabrafenib+cetuximab. As illustrated on Figure 8b, tumours with mutations of MAP2K1 I111 to T and / or Q110 to H and / or I112 to T may be treated with a therapy that is not trametinib or dabrafenib+cetuximab. 64 The present examples show for the first time, an association between: (i) mutation of MAP2K1 L115 to P and (ii) resistance to trametinib. As illustrated on Figure 8b, tumours with mutations of MAP2K1 L115 to P may be treated with a therapy that is not trametinib, such as e.g. dabrafenib+cetuximab. The present examples show for the first time, an association between: (i) mutation of MAP2K1 H119 to R 5 and / or E120 to G and (ii) resistance to each of trametinib, dabrafenib+cetuximab and adagrasib. Data for at least of trametinib and dabrafenib+cetuximab indicates that tumours with these variants are likely to have drug addiction to these drugs. As illustrated on Figure 8b, tumours with mutations of MAP2K1 H119 to R and / or E120 to G may be treated with a therapy that is not trametinib, dabrafenib+cetuximab or adagrasib, or may be treated with an intermittent schedule of one or more of trametinib, dabrafenib+cetuximab and 10 adagrasib. The present examples show for the first time, an association between: (i) mutation of MAP2K1 N122 to S or D and / or E120 to G and (ii) resistance to each of trametinib, dabrafenib+cetuximab, adagrasib and sotorasib. Data for at least of trametinib and dabrafenib+cetuximab indicates that tumours with these variants are likely to have drug addiction to these drugs. Data for sotorasib and adagrasib indicates that 15 tumours with these variants are likely to have canonical drug resistance to these drugs. As illustrated on Figure 8b, tumours with mutations of MAP2K1 N122 to S or D and / or E120 to G may be treated with a therapy that is not trametinib, dabrafenib+cetuximab, adagrasib or sotorasib, or may be treated with an intermittent schedule of one or more of trametinib and dabrafenib+cetuximab. The present examples show for the first time, an association between: (i) mutation of MAP2K1 Y130 to C 20 or H and (ii) resistance to each of trametinib and dabrafenib+cetuximab. Data for trametinib and dabrafenib+cetuximab indicates that tumours with these variants are likely to have drug addiction to these drugs. As illustrated on Figure 8b, tumours with mutations of MAP2K1 Y130 to C or H may be treated with a therapy that is not trametinib or dabrafenib+cetuximab, or may be treated with an intermittent schedule of one or more of trametinib and dabrafenib+cetuximab. 25 The present examples show for the first time, an association between: (i) mutation of MAP2K1 Y130 to C and (ii) resistance to each of sotorasib and adagrasib. Data for sotorasib and adagrasib indicates that tumours with these variants are likely to have canonical drug resistance to these drugs. Tumours with mutations of MAP2K1 Y130 to C may be treated with a therapy that is not sotorasib or adagrasib. The present examples show for the first time, an association between: (i) mutation of MAP2K1 S194 to P 30 and (ii) resistance to each of trametinib, sotorasib, adagrasib and dabrafenib+cetuximab, Data for at least trametinib and dabrafenib+cetuximab indicates that tumours with these variants are likely to have drug addiction to these drugs. As illustrated on Figure 8b, tumours with mutations of S194 to P may be treated with a therapy that is not trametinib, sotorasib, adagrasib or dabrafenib+cetuximab, or may be treated with an intermittent schedule of one or more of trametinib, sotorasib, adagrasib and dabrafenib+cetuximab. 35 The present examples show for the first time, an association between: (i) mutation of MAP2K1 E203 to K and / or D208 to N and (ii) resistance to each of trametinib and dabrafenib+cetuximab. As illustrated on 65 Figure 8b, tumours with mutations of MAP2K1 E203 to K and / or D208 to N may be treated with a therapy that is not trametinib or dabrafenib+cetuximab. The present examples show for the first time, an association between: (i) mutation of MAP2K1 V211 to A and (ii) resistance to trametinib. As illustrated on Figure 8b, tumours with mutations of MAP2K1 V211 to A 5 may be treated with a therapy that is not trametinib, such as e.g. dabrafenib+cetuximab. Methods Cell lines. All cell models used in this study (H23, PC9, HT-29, MHH-ES-1, HEK293T) were verified as mycoplasma-free and STR profiled in accordance with authentication guidelines (Zhu 2019). Cells werecultured in RPMI medium with 10 % FCS and 1X penicillin-streptomycin (Thermo Fisher Scientific) at 3710 °C in a humidified incubator with 95 % air and 5 % CO2. gRNA library design and generation. To generate base editing gRNA controls the inventors used SpliceR (Komor 2016) and ranked the top three gRNAs using the following metric [cDNA disruption score x (ABEscore + CBEscore)]. For exonic gRNA designs the inventors used BEstimate (manuscript in preparation; github.com / CansuDincer / BEstimate). gRNAs were designed against the reference genome 15 GRCh38, and their design did not integrate cell line-specific SNVs. Oligo pools (Twist Biosciences) were PCR amplified and inserted into a Bbs-I digested pKLV2-BFP-T2A-puroR lentiviral backbone (Addgene plasmid #67974) using Gibson assembly (NEB). Libraries from Gibson assembly were ethanol precipitated before delivery into electrocompetent cells (Endura, Lucigen) by electroporation with multiple parallel transformations to maintain library representation, before propagation in LB supplemented with 100 µg / ml 20 ampicillin, shaking at 30 °C overnight. For virus packaging, HEK293T cells were co-transfected with the library plasmid pool, psPAX2 and pMD2.G plasmids using FuGene HD (Promega) in Opti-MEM (Thermo Fisher Scientific). Viral particles were collected in media supernatant 72 h post-transfection, filtered and frozen. For the perturb-seq library, the inventors selected gRNAs that were top resistance hits for different drugs 25 in the base editing screens. The inventors introduced an iBAR barcode into the gRNA stem-loop (Zhu 2019), by using a primer with a hexanucleotide degenerate sequence (Sigma-Aldrich) to amplify the oligo library (Twist Biosciences). This enables the identification of genetically identical clonal populations of daughter cells expressing the same gRNA. Base editing screens. Base editing cell lines were generated as previously described (Coelho 2023). Briefly, 30 the inventors integrated doxycycline-inducible CBE or ABE NGN base editors into the CLYBL safe-harbour locus using CRISPR-Cas9 and homology-directed repair, and used blasticidin and mApple (FACS) to select for base editing cell populations. The inventors used BE-FLARE (Coelho 2018) (CBE) or a GFP stop codon reporter (Fu 2021) (ABE) to measure overall base editing activity. For MHH-ES-1, the inventors only performed ABE mutagenesis screens due to the poor performance of CBE in this cell model. 35 The inventors transduced cells with the gRNA virus (plus 8 µg / ml polybrene; Thermo Fisher Scientific) to achieve an infection rate of ~30 %, as measured by BFP fluorescence with flow cytometry. The inventors 66 performed all base editing screens at ~1000 X coverage (estimated cells / gRNA). Cells were selected with puromycin (Thermo Fisher Scientific) for four days before taking a time 0 cell pellet and then induction of base editing with doxycycline (1 µg / ml; Sigma Aldrich) for three days. Screens were split into drug treatment arms or control arms for a further 11 days, with passaging or refreshing drug every 3-4 days. Drug 5 concentrations were based on IC50 from GDSC (Yang 2013) and empirically confirmed using Cell Titre Glo proliferation drug titration assays. Drug concentrations used in screens and validation experiments were as follows: trametinib 10 nM, pictilisib 1 µM, dabrafenib 80 nM, cetuximab 1 µg / ml, sotorasib 2 µM, adagrasib 1 µM, gefitinib 75 nM, osimertinib 5 nM, olaparib 510 nM, niraparib 330 nM (Selleckchem). Each screen was independently repeated at least twice on separate weeks. 10 DNA was extracted from cell pellets (Qiagen) and the gRNA cassette was PCR amplified (PCR1, 28 cycles, 3 µg / reaction) with multiple reactions in parallel to maintain the complexity of the library. After column purification of PCR1 products (Qiagen), PCR products were indexed by PCR (PCR2, 8 cycles, 0.2 ng / reaction). Libraries were sequenced on the HiSeq2500 (Illumina) using 19 bp SE sequencing on Rapid Run mode with a custom primer. All gRNA validation primers are listed in Table 6 below. 15 Screen analysis. For analysis of predicted edits and annotation of gRNAs, the inventors used BEstimate (github.com / CansuDincer / BEstimate). To infer the effect of edits, the inventors assumed a WT genome and did not consider SNPs or SNVs specific to the four cancer cell lines. To score gRNAs, the inventors first normalised read counts by generating reads per million (RPM) [reads / total reads for sample *1,000,000 + 1 pseudo-count]. For replicate screens, the inventors averaged the read count values of the same gRNAs 20 to generate an average RPM. The inventors then generated average log2fold-change (L2FC) values as [log2(RPM condition / RPM control)]. The z-scores were generated as [(L2FC – mean L2FC) / sd], where sd represents the standard deviation of the L2FC values for that comparison. The inventors excluded gRNAs that had more than two perfect matches in the GRCh38 human genome (n=134), and gRNAs with < 100 read counts in the plasmid or any time 0 sample in the screens (n=27). For the few hit gRNAs with 25 two perfect matches in the genome, the inventors confirmed there was only one exonic target. The inventors also excluded gRNAs specifically in the MHH-ES-1 screen that had a 10-fold read count difference between replicates (n=118) from downstream analysis of the MHH-ES-1 screens. The inventors assigned drug resistance hits as gRNAs with a z-score >2 in the presence of drug, and >1 in the presence of drug for each independent replicate. For variant classes based on proliferation screens, 30 the inventors assigned driver variants to gRNAs with a z-score >2 in the absence of drug, and >1 in each independent replicate. Drug addiction variants were assigned to gRNAs with a z-score <-2 in the absence of drug (and <-1 in each independent replicate), and >2 in the presence of drug (and >1 in each independent replicate). Canonical drug resistance variants were assigned to gRNAs with a z-score >2 in the presence of drug (and >1 in each independent replicate) and did not fulfil driver or drug addiction phenotypic criteria. 35 gRNAs with average z-scores <-2 in the presence of drug (and <-1 in each replicate), and >-2 in the absence of drug (and >-1 in each replicate), were assigned as drug-sensitising hits. Control gRNAs did not satisfy any of the above criteria. Boolean classification of drug resistance irrespective of variant class was 67 defined as z-score >2 in the presence of drug (and >1 in each independent replicate). For osimertinib and gefitinib resistance hits, the PC9 single-cell validation library screens were also used to further verify resistance (further threshold of z-score of > 2 in these screens). The inventors used MAGeCK (Li 2014) to generate statistics for base editing and prime editing screens. For base editing screens the inventors used 5 the control gRNAs (non-targeting, intergenic and non-essential) to generate a control distribution and implemented MAGeCK (RRA) to compare against the test gRNAs. For prime editing screens, the inventors used the same approach but the control distribution was generated from pegRNAs designed to install synonymous variants. All gRNA and pegRNAs hits had MAGeCK-p-value <0.05 and an FDR <0.1. For COSMIC variants, the inventors downloaded non-synonymous variants in our 11 target genes (February 10 2024) that pertained to resistance to inhibitors used in this study or similar drugs targeting the same oncogenes. This included patient samples and patient-derived xenograft in vivo samples with post- treatment biopsy genotyping of variants and an associated DRUG_RESPONSE field. Prime editing. MLH1 was knocked-out (KO) in PC9 cells with CRISPR / Cas9 using a Cas9-T2A-EGFP (Addgene plasmid #48140) plasmid using a gRNA with sequence GCACATCGAGAGCAAGCTCC (SEQ 15 ID NO: 83). Cells were transfected (FuGene HD, Promega), sorted on EGFP by FACS as single cells into 96 well plates (BD Influx, BD Biosciences), and clones were tested for KO by Western blotting (see below) with an MLH1 antibody (#3515, Cell Signaling Technologies). MLH1 KO PC9 clone 1, or parental PC9 cells were co-transfected with a PiggyBac doxycycline-inducible PE2 plasmid with a constitutive GFP selection marker and PiggyBac transposon plasmid (Koeppel 2023) at a 1:1 ratio. GFP positive cells were sorted by 20 FACS as a pooled population of PE-expressing cells. EGFR C797S pegRNAs are listed. PegRNAs were designed by taking the top 3 scoring pegRNAs as predicted by a recent pegRNA efficiency prediction algorithm, PRIDICT (Mathis 2023). The inventors used an end-to-end batching algorithm to call the PRIDICT tool for multiple query amino acids github.com / mariemoullet / PRIDICT. The inventors generated lentiviral pegRNAs (Twist Biosciences or IDT) as described above for base editing libraries, 25 except they used an epegRNA acceptor puromycinR plasmid using a BsmBI entry site for the gRNA (Koeppel 2023) (modified from Addgene plasmid #84752). After selection with puromycin, prime editing was performed by 5-7 days exposure to doxycycline (1 µg / ml). Prime editing screens. For prime editing, the inventors selected coding positions of EGFR that scored in the base editing screens for EGFR inhibitors, and selected all possible amino acid mutations possible with 30 an SNV for the following residues A1013, C797, T790, D1006, D1012, V1011. The inventors designed the pegRNAs using PRIDICT as above and filtered pegRNA inserts that were <200 bp in length after the addition of Gibson homology arms for cost-effective DNA oligo synthesis (Twist Biosciences). A library of 162 vectors (Table 2, SEQ ID NO: 147-308) was cloned using Gibson assembly (NEB), to generate a pool of plasmids for lentivirus production, as described above for base editing screens. PC9 MLH1 KO PE2- 35 GFP expressing cells were infected at such that approximately 15-30 % of cells were infected and to achieve a library coverage of approximately 10,000 X, and cells were subsequently selected with puromycin for 3 days (0.75 µg / ml). Prime editing was induced for 7 days with doxycycline (1 µg / ml), before selection 68 with osimertinib (75 nM) for 10 days, or grown in DMSO (control). Cells were pelleted, washed in PBS and DNA was extracted (Qiagen) for pegRNA amplification using primers (Table 2, SEQ ID NO: 84-85). Amplicons were sequenced with paired-end sequencing on a MiSeq v2 micro kit (Illumina). For validation studies, the inventors used an epegRNA design of pegRNAs that had scored in the prime 5 editing screens (Table 2, SEQ ID NO: 86-90). A NT gRNA vector expressing GFP was used as a WT control in co-competition assays against epegRNAs expressing BFP (including a EGFR C797C synonymous control), which were analysed by flow cytometry five days after seeding. Ratios of BFP to GFP were normalised to the day 0 timepoint. Senescence assays. Oncogene-induced senescence was assessed by staining with the ß-galactosidase10 staining kit 48 h after drug treatment in 96 well plates following manufacturer’s instructions (Cell Signaling Technology, #9860) and images were acquitted on the EVOS XL Core microscope (Thermo Fisher Scientific). Proliferation assays. For validation assays the inventors used the incucyte (Sartorius) or CellTiter-Glo assay (Promega). gRNAs were ordered as oligos (Sigma-Aldrich), annealed and assembled by Golden Gate and 15 sequence verified (Eurofins), as described (Coelho 2023). This allowed the inventors to validate gRNAs at scale in an arrayed format. gRNA sequences are provided above (SEQ ID NO: 1-82). Western blotting. PC9 cells were lysed in sample loading buffer (8% SDS, 20 % b-mercaptoethanol, 40 % glycerol, 0.01 % bromophenol blue, 0.2 M Tris-HCL pH 6.8) and boiled for 5 minutes before loading onto a NuPAGE 4-12 % Bis-Tris gel (Thermo Fisher Scientific). Proteins were transferred to a PVDF membrane 20 before blotting with the following primary antibodies: EGFR total (1068 epitope, #2232), p-EGFR (Tyr1148, #4404), b-actin (#4970), p-ERK (#9101), ERK total (#9102) (Cell Signaling Technology), EGFR (epitope 1020-1046) #610017 (BD Biosciences). Secondary antibodies were conjugated to horseradish peroxidase. Flow cytometry. PC9 cells were harvested by trypsinisation and washed in FACS buffer (0.5 % FCS< 2 mM EDTA in PBS) before staining with anti-EGFR-FITC antibody (#MA5-28104, Thermo Fisher Scientific) 25 for 25 min on ice in the dark. For B2M (#395805, BioLegend) and HLA staining (#MA5-44095, Thermo Fisher Scientific), cells were treated the indicated drug for 48 h before analysis. Cells were washed twice in FACS buffer, incubated with DAPI (1 µg / ml, Thermo Fisher Scientific) before filtering through a nylon mesh cell strainer, and analysis on an LSRFortsessa (BD Biosciences). RNA analysis. RNA was extracted (RNeasy, Qiagen), DNA removed with DNaseI digest (Qiagen), and 30 reverse transcription performed using poly dT priming (SuperScript IV, Thermo Fisher Scientific). PCR was performed on cDNA from WT and mutant cells and intron retention was verified by the size of the resulting PCR product as assessed by DNA gel electrophoresis. EGFR primer sequences: forward TTGACTGAGGACAGCATAGACGAC (SEQ ID NO: 91) / reverse: GGTTCAGAGGCTGATTGTGATAGAC (SEQ ID NO: 92). 35 Amplicon sequencing. From the drug resistance hits from base editing screens (z-score of >2 and >1 in each replicate with FDR <0.1 and p-value <0.05), the inventors filtered for non-synonymous coding 69 mutations in target genes that were non-redundant and had not already been genotyped by Sanger sequencing in validation studies (e.g. EGFR splice variants and PARP1 drug-sensitising variants). Given the large number of variants, the inventors preferentially selected proximal variants such that we could cover multiple variants in a single amplicon. In total, the inventors analysed 45 gRNAs targeting 7 genes 5over 21 amplicons in two separate experiments. HT-29 cells were directly lysed with a direct PCR lysisbuffer supplemented with 100 µg / ml proteinase K following manufacturer’s instructions (Viagen Biotech).2 µl of DNA lysate was used as input in 25 µl PCR reactions (KAPA HiFi HotStart ReadyMix, Roche) using amplicon sequencing primers listed in Table 3 (SEQ ID NO: 93-95, 101-146) PCR cycle number for PCR1 was determined empirically. PCR products were SPRI purified (AMPure XP, 10 Beckman Coulter), before indexing PCR2 (10 cycles), SPRI purification, quantification (Bioanalyzer, Agilent), equimolar pooling, and sequencing on a MiSeq v2 150 bp PE kit (Illumina). Similarly, Sanger sequencing (Eurofins) of base edits were performed by PCR amplification of endogenous edited loci using primers listed in Table 3. For analysis, the inventors used BCFtools and a vafCorrect, with a read-depth cut-off of >1000 reads, and a VAF > 0.1 for variants. The editing outcomes from amplicon sequencing are 15 reported in Table 5C. 43 gRNAs produced non-synonymous variants with >10% variant allele frequency and were considered in downstream analysis. For the annotation of MAP2K2 Y134H, the inventors deduced this was the variant driving drug resistance as they observed two different gRNAs making more complex edits with Y134 as the common major allele (97 % median allele frequency for one gRNA). The inventors removed the MAP2K2 F213C variant from downstream analysis as it was already detected in unedited 20 DNA samples. Perturb-seq . After infection with the barcoded validation gRNA library virus to achieve an infection rate of ~20 %, HT-29 cells were selected with puromycin and then the cell population was bottlenecked to 50,000 cells to increase the numbers of genetic clones. CBE and ABE base editing was induced for 48 h with the addition of doxycycline (1 µg / ml). Base edited cells were then maintained in puromycin (0.5 µg / ml) to 25 maintain gRNA expression, and acutely treated with doxycycline (to express base editor and help stabilise gRNAs) and dabrafenib and cetuximab for 16 h before harvesting the cells for transcriptomics. Single cell suspensions of 60 K cells per reaction were prepared for super-loading on the Chromium X according to manufacturer’s instructions using Chip N and the Chromium Next GEM 5’ HT v2 kit (10X Genomics). The inventors spiked in a 2.2 µL of a 10 µM of a primer for direct capture of the gRNA in the RT reaction. The30 gRNA libraries were prepared separately using a nested PCR. Primers are listed in Table 7 below. The cDNA gene expression libraries were prepared according to manufacturer’s instructions (10X Genomics), before pooling at 1:10 molar ratio with the gRNA libraries (gRNA:cDNA) and sequencing on one lane of a NovaSeq 6000 S4 (Illumina). Perturb-seq analysis - Processing and quality control. The inventors used Cell Ranger 7.0.1 to obtain UMI 35 counts for gRNA and mRNA and for cell-calling. For quality control, the inventors removed low outliers for the total count, low outliers for the number of detected features and high outliers for the percentage of counts from mitochondrial genes using the scater (McCarthy 2017) Bioconductor package, obtaining 70 56,220 cells (ABE data set: 28,718; CBE data set: 27,502) out of 58,673 cells called by Cell Ranger in the first instance. Perturb-seq analysis - gRNA calling. The inventors used Cell Ranger to obtain UMI counts for all gRNAs and used a robust probabilistic mixture modelling method to call gRNAs that they had developed previously 5 (Cooper 2024) to distinguish between higher UMI counts that correspond to a gRNA being in a cell, versus ambient background counts. The inventors defined two thresholds for UMI counts; a lower threshold – UMI counts below this threshold mean a 90% probability of being ambient noise; and an upper threshold-UMI counts above this threshold are ambient noise with a probability of 10%. gRNAs were called in a cell if their UMI counts were above the upper threshold and no other gRNA had UMI counts between the lower and 10 upper thresholds. Excluding any cells with splice-essential positive control gRNAs, the inventors’ robust gRNA calling resulted in 27,823 cells with confident gRNA assignment (ABE data set: 15,496, CBE data set: 12,327). The average number of cells per gRNA excluding the splice essential gRNAs is 86 for ABE and 63 for CBE, where cells with multiple gRNAs are counted towards each of those gRNAs. iBAR calling and iBAR groups. First, the inventors used Cell Ranger to obtain UMI counts for all possible 15 iBAR barcodes. Then, the inventors assigned iBAR barcodes to all cells with gRNAs, by assigning the iBARs with the n highest UMI counts to a cell with n gRNAs. For downstream analysis, the inventors combined cells with the same gRNA and the same iBAR to iBAR-groups, by combining their counts and then normalising for differences in cell sizes and numbers of aggregated cells by means of library normalisation using the scuttle Bioconductor package. iBAR-groups are therefore small groups of cells with 20 identical genotypes. Averaging across the iBAR-groups with more than one cell and treating each iBAR group like a cell in downstream analysis avoids biases in analysis such as false positives or negatives for differential expression (DE) analysis resulting from different numbers of cells with different edits. The use of similar cellular identifiers has been shown to improve accuracy in pooled CRISPR-screens (Zhu 2019, Michlits 2017). For general scRNA-seq analysis, unmodelled heterogeneity has been identified as a major 25 cause of false positives in differential expression analysis (Squair 2021). For the inventors’ screen, unlike traditional perturb-seq screens, they were therefore able to account for differences in editing and the clonal structure of the data using the iBARs. gRNA-level meta-data from the pooled experiments such as variant type (NT controls, canonical drug resistance, driver, drug addiction) or whether the gRNA confers drug resistance were assigned to iBAR 30 groups as follows: if the iBAR group has only one gRNA, then the data from this gRNA was assigned. In the case of several gRNAs, the stronger impact variant type was assigned (driver / drug addiction > canonical drug resistance > control, and iBAR-groups with both driver and drug addiction variant gRNAs were discarded from further analysis). The inventors obtained 9,505 iBAR-groups for ABE and 8,171 for CBE. The average number of iBAR groups per gRNA was 56 for ABE and 45 for CBE, including iBAR groups 35 with multiple gRNAs / iBARs. The mean number of cells per iBAR-group was 1.6 for ABE and 1.5 for CBE. Pathway analysis . The inventors performed pathway analysis using PROGENy for the pathways available with the tool (Schubert 2018). PROGENy scores are based on a large collection of external publicly 71 available perturbation experiments. The inventors computed a separate PROGENy score for each iBAR- group, and mean scores per gRNA. To improve comparability of scores across pathways, the inventors transformed the scores linearly such that their mean on the NT gRNAs is 0 and the standard deviation across the same barcodes is equal to 1. They also tested for each of the gRNAs identified in the pooled 5 screens as conferring canonical drug resistance, driver variants, or drug addiction, whether there is significant differential expression of pathway scores between the respective gRNA and NT controls, using the same test described for gene expression (Fig.13c and Fig.2-5). To extend the number of pathways considered, the inventors also used MAYA (Landais 2023) with the MSigDB Hallmark pathways (Liberzon 2015) as input gene lists. For each pathway, MAYA performs PCA 10 on the input data set restricted to the genes of the pathway and selects those pathways and principal components for which there is bimodality in the data. The inventors then used the pathways flagged as bimodal by MAYA to perform analysis analogously to that using PROGENy scores (Fig.6-7). Differential gene expression analysis. The inventors performed differential gene expression (DE) analysis tests at the gRNA level for all gRNAs associated with drug resistance, drug addiction or driver variants,15 which are the only gRNA (as opposed to one of multiple gRNAs in an iBAR-group) for at least 10 iBAR- groups (Fig. 2-3). The inventors obtained 9,505 iBAR cell groups for ABE and 8,171 for CBE, with an average of 56 iBAR groups per gRNA for ABE and 45 for CBE. iBAR groups were treated as a unit in downstream analysis, with their gene expression averaged across the cells forming the group. The DE tests were performed by comparing iBAR-groups containing each test gRNA (as unique gRNA in the iBAR-group 20 or one of several gRNAs) to the iBAR-groups with non-targeting gRNAs, using the non-parametric Wilcoxon rank sum test, since the iBAR-groups cannot be assumed to follow a parametric distribution like a Gaussian or a negative Binomial distribution, as they may have been impacted to different degrees by the gRNAs, and editing may have been successful in some gRNA-groups and unsuccessful in others (Cooper 2024). The inventors also tested for each of the gRNAs identified in the pooled screens as conferring 25 canonical drug resistance, driver mutations, or drug addiction, whether there was significant differential expression of PROGENy or MAYA pathway scores between the respective gRNA and NT controls. (Fig.2- 3 and Fig. 21c). The inventors selected the impacted gRNAs with at least one gene or pathway with a p- value less than 10^-6, as a measure of a minimum response to the perturbation (13 (of 20) gRNAs for the ABE and 11 (of 17) for the CBE data set). The inventors controlled the expected average false discovery 30 rate for DE testing for the selected gRNAs as described in Benjamini & Bogomolov 2014. The number of significantly differentially expressed genes for the selected gRNAs at expected average FDR of 0.1 ranged from 28 to 2624. The inventors also performed DE tests at the level of variant class, comparing drug addiction to canonical drug resistance. Here, the inventors subsampled the same number of iBAR groups for each gRNA to obtain 35 a general result unbiased by specific gRNAs or target genes, using the standard Benjamini-Hochberg correction of false discovery rate across genes and pathway scores (Benjamini & Hochberg, 1995). 72 Dimensionality reduction, clustering and cell-cycle analysis. Principal component analysis (PCA) was performed at the iBAR-group level using genes identified as DE at FDR<0.1 for at least one test gRNA and with a log2 fold-change of more than 0.25. A UMAP representation was computed based on this PCA. Clustering was performed using the Louvain method. Significance of differences between proportions of 5 variant types was tested for using the standard chi-squared test. Scores for cell-cycle phases were computed using Seurat (Butler 2018). DE analysis across variant types. At both the levels of genes and pathway scores, the inventors performed DE analysis between different types of variants, e.g. between drug addiction and canonical drug resistance. To avoid biases resulting from different abundance of different gRNAs within the variant types, the inventors 10 subsampled the data such that each gRNA contributes at most 20 iBAR-groups. Diffusion scores, energy distance and PFS scores. The inventors computed diffusion scores as described in their previous single-cell CRISPR screen (Cooper 2024), for all gRNAs associated with drug addiction or resistance (ABE) with at least 10 iBAR groups with the specific gRNA as the only gRNA in the cell. The diffusion scores order the iBAR-groups in terms of their progression towards the strongest effect driver / drug 15 addiction variants. Diffusion scores were transferred to gRNAs associated with canonical drug resistance from the CBE dataset by mapping each iBAR group from CBE to the one most similar in the ABE dataset, where similarity was measured in terms of rank correlation on the genes most rank correlated with the diffusion score in the ABE dataset. Then the CBE iBAR group was assigned the mean diffusion score of those five ABE iBAR groups. Clusters of gRNAs were transferred similarly. Each CBE iBAR group was 20 assigned the most frequent cluster among the cluster associations of the five most rank-correlated ABE iBAR groups. A number of gRNAs had bimodal diffusion scores, with a cluster of iBAR groups with higher and one with lower diffusion scores. These were identified by applying a mixture of normal distributions with identical variances, using the mclust R package (Scrucca 2016) to identify the best number of clusters from between 25 1 and 5 clusters, with an optimal cluster number of 2 interpreted as bimodality. Energy distances (Replogle 2022) were computed between targeting and NT gRNAs for all gRNAs that are the only gRNA in the cell for at least 10 iBAR-groups, based on the principal components obtained from all genes differentially expressed for at least one gRNA for the ABE or CBE data set and using iBAR-groups rather than cells. For comparability between the ABE and CBE data sets, we computed for each data set the median energy 30 distance across non-targeting gRNAs, and divided energy distances for targeting gRNAs by that number for the respective data set. The inventors used published differential gene expression results for BRAF- V600E colorectal cancer patients treated with a combination of PD-1, BRAF and MEK inhibition (Tian 2023). Tian et al. reported log2-fold gene expression changes at 15 days versus pre-treatment bulked across patients with progression free survival (PFS) > 6 months, and bulked across those with PFS < 6 months. 35 To perform correlation analysis to the data, the inventors identified the genes with an absolute log2-fold expression change of more than 0.25 and adjusted p-value of less than 0.1 for either the data for PFS > 6 months or PFS < 6 months (2148 genes). Then, based on this subset of genes, the inventors used 73 Spearman rank correlation to correlate the log2-fold expression changes for PFS > 6 months reported in Tian et al with the log2-fold changes from differential expression testing results herein for individual targeting gRNAs compared to non-targeting controls (PFS outcome scores). Tumour organoid T cell co-culture assays. Tumour organoid co-culture assay were performed as described 5 (Coelho 2023, Cattaneo 2020, Dijkstra 2018), except the tumour cells were not pre-treated with IFN-gamma as the proficiency of antigen presentation was being tested. For co-competition assays, BFP gRNA tumour cells were co-cultured with NT gRNA-GFP-expressing cells in a co-competition assay. Briefly, PMBCs were cultures in 96 well-plates coated with anti-CD28 antibody for 24 h with IL-2 (150U / ml; Thermo Fisher Scientific). Autologous CRC-9 tumour organoids were maintained in 80 % basement membrane extract 10 (R&D systems), pre-treated for 48 in the MEK inhibitor trametinib (25 nM), and plated (5,000 cells of BFP gRNA and 5,000 cells of GFP gRNA) in suspension at a 3:1 E:T ratio for 72 h before FACS analysis. T cell killing assays were performed in the presence of the anti-PD1 antibody nivolumab (20 µg / ml; Selleckchem) in RPMI medium supplemented with human serum and primocin (Invivogen). 123count eBead counting beads (Thermo Fisher Scientific) were used to calculate absolute cell counts from flow cytometry analysis. 15 Relative T cell killing was quantified by normalising the number of cancer cells in comparable conditions in the absence of T cells. Data availability. Sequencing data are deposited on ENA and accessions are listed in Table 4. Genotyped variants and processed data including screen z-scores will be released to the MAVE database (Esposito 2019) upon publication. 20 Code availability. Code used to analyse base editing screens can be found on GitHub here: github.com / MatthewACoelho / Res1_analysis. Code used to analyse the single-cell screens can be found here: github.com / MarioniLab / BE_perturb_seq_drug_resistance. Tables Table 1 – List of target cancer genes, locations, and corresponding Ensembl IDs for gene, transcript, 25 and exon. Coordinates provided are those in reference genome GRCh38. 74 75 76 77 78 79 80 81 Table 2 – pegRNAs 82 Table 3 – Primers for amplicon sequencing 83 Table 4 – ENA (European Nucleotide Archive) accession numbers - www.ebi.ac.uk / ena ibrapairn 84 85 86 87 88 89 90 91 92 93 Table 5A. Description of COSMIC-curated drug resistant variants with post-treatment DRUG_RESPONSE fields for relevant drugs and gene targets in this study. Non-synonymous variants are shown. Highlighted in bold are cases where at least one base editing gRNA is in agreement with the clinical patient response or relevant xenograft / PDX drug response in vivo. Underlined are variants not 94 covered by any gRNA in base editing screens. T=TRUE, F=FALSE, N=NA (indicates whether or not there was resistance to the drug indicated). One row per individual guide RNA associated with the mutation. 95 AF F T T F F F F F A8;9;10;9;10 E;N / D;N / S; m F F F T F F F F F c c c c c c c c cC Q / R;ENQ / Y M EGR A63 K / R;K / E;K / m F F F T F F F F F c c c c dr c c c cG A103 I / T m F F F F T F F F F c c c c c c c c drA 134 Y / C m F F T T F F F F F dr c c c dr c c c cA2 K2 137 F / L m F F T T F F F F F dr c c c dr c c c cCPA M149 H / Y m F F T F F F F N N c c c c c c N N cA 198 S / P m F F T T F F F F F dr c c c dr c c c cA 198 S / P m F F T T F F F F F c c c c c c c c cA 198 S / P m F F T T F F F F F dr c c c dr c c c cA 198 S / P m F F T T F F F F F dr c c c dr c c c c 96 A2 I T F F T T F F F F F A50;49;48;4Q / R;E / G;Em F F T F F F F F F c c c c c c c c c8;50 QQ / GRR A50;49;49;5Q / R;QQ / Rm F F T T F F F F F dr c c c dr c c c c0 R A51;50;50;5K / R;K / E;Q / m F F T T F F F F F dr c c c dr c c c c1 R;QK / RG A31 E / G m F F T F F F F F F c c c c c c c c cA 96 R / G m F F T F F F F F F c c c c c c c c cA1111 I / T m F F F F T T F F F c c c c c c c c cAK2P111 I / T m F F F F T F F F F c c c c c c c c cA A M111 I / T m F F F F T F F F F c c c c c c c c cA 194 S / P m F F T T T T F F F c c d c c c c c da a 97 98 99 100 AP F F T T T T F F F C325 D / N m F F F F F F T N N c c c c c c N N cC 275;274 L / F;D;DL / m F F T F F F F N N c c c c c c N N cDF C 1323;322;32D / N;E;E / K; Tm F N F F F F T N N c c d d c c N N dK A 1;322;323 L;LED / LK N C323;322;32D / N;E;E / K;m F F F F F F T N N c c d d c c N N d2;323 ED / KN A331;329 D / G;A;AV m F F F F T F F F F c c c c c c c c cD / AVG C378;376;37S / F;A;A / V;m F F F F F F T N N c c c c c c N N c6;378 AKS / VKF 101 Table 5B. gRNAs conferring a drug resistant phenotpye to at least one drug with no associated DRUG_RESPONSE data from COSMIC are listed. Predicted amino acid substitutions are shown. T=TRUE, F=FALSE, N=NA (indicates whether or not there was resistance to the drug indicated). C=CBE, A=ABE (editor). c=control, dr=canonical drug resistance, da=drug addiction. ds=drug-sensitsing. 5 m=missense, sc=stop codon, sl=start lost, sv=splice variant. Experimentally validated dRNAs are in bold. The genotyped edit from amplicon sequencing is given in Table 5C if available. gRNAs fpr which genotype from amplicon seq is available are indicated by the use of italics and bold font in the “editor” column. 102 Table 5C. gRNAs conferring a drug resistant phenotype to at least one drug with no associated DRUG_RESPONSE data from COSMIC – amplicon seq data. 103 104 105 106 107 108 109 110 111 112 113 Table 5D. Genotyping data from NGS amplicon sequencing of endogenous base edits for hit gRNAs. 114 115 Table 6 – gRNA validation primers Table 7. Primers used in perturb-seq experiments 116 117 118 119 120 121 Table 8. Genotyping data from NGS amplicon sequencing of endogenous base edits for hit gRNAs 122 123 124 125 Table 9. Novel variants – drug response associations. 126 Table 9 (continued). Novel variants - Amplicon seq edits where available. References A number of publications are cited above in order to more fully describe and disclose the invention and the state of the art to which the invention pertains. Full citations for these references are provided below. The 5 entirety of each of these references is incorporated herein. 1. Vasan, N., Baselga, J. & Hyman, D.M. A view on drug resistance in cancer. Nature 575, 299-309(2019). 2. Pao, W. et al. Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is associated with a second mutation in the EGFR kinase domain. PLoS Med 2, e73 (2005).10 3. van de Haar, J. et al. Limited evolution of the actionable metastatic cancer genome under therapeutic pressure. Nat Med 27, 1553-1563 (2021).4. Cuella-Martin, R. et al. Functional interrogation of DNA damage response variants with base editing screens. Cell 184, 1081-1097 e19 (2021).5. Hanna, R.E. et al. Massively parallel assessment of human variants with base editor screens. Cell 15 184, 1064-1080 e20 (2021). 6. Kim, Y. et al. High-throughput functional evaluation of human cancer-associated mutations using base editors. Nat Biotechnol 40, 874-884 (2022).7. Sanchez-Rivera, F.J. et al. Base editing sensor libraries for high-throughput engineering and functional analysis of cancer-associated single nucleotide variants. Nat Biotechnol 40, 862-873 (2022).20 8. Coelho, M.A. et al. Base editing screens map mutations affecting interferon-gamma signaling in cancer. Cancer Cell (2023). 9. Lue, N.Z. et al. Base editor scanning charts the DNMT3A activity landscape. Nat Chem Biol 19,176-186 (2023). 10. Lue, N.Z. & Liau, B.B. Base editor screens for in situ mutational scanning at scale. Mol Cell 83,25 2167-2187 (2023). 11. Martin-Rufino, J.D. et al. Massively parallel base editing to map variant effects in human hematopoiesis. Cell 186, 2456-2474 e24 (2023).12. Cooper, S., Obolenski, S., Waters, A.J., Bassett, A.R. & Coelho, M.A. Analyzing the functional effects of DNA variants with gene editing. Cell Reports Methods, 100776 (2024). 30 13. Sangree, A.K. et al. Benchmarking of SpCas9 variants enables deeper base editor screens of BRCA1 and BCL2. Nat Commun 13, 1318 (2022).14. Bock, C. et al. High-content CRISPR screening. Nat Rev Methods Primers 2(2022). 127 15. Anzalone, A.V., Koblan, L.W. & Liu, D.R. Genome editing with CRISPR-Cas nucleases, base editors, transposases and prime editors. Nat Biotechnol 38, 824-844 (2020).16. Komor, A.C., Kim, Y.B., Packer, M.S., Zuris, J.A. & Liu, D.R. Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage. Nature 533, 420-4 (2016).5 17. Gaudelli, N.M. et al. Programmable base editing of A*T to G*C in genomic DNA without DNA cleavage. Nature 551, 464-471 (2017).18. Xu, P. et al. Genome-wide interrogation of gene functions through base editor screens empowered by barcoded sgRNAs. Nat Biotechnol 39, 1403-1413 (2021).19. Yang, W. et al. Genomics of Drug Sensitivity in Cancer (GDSC): a resource for therapeutic10 biomarker discovery in cancer cells. Nucleic Acids Res 41, D955-61 (2013).20. van der Meer, D. et al. Cell Model Passports-a hub for clinical, genetic and functional datasets of preclinical cancer models. Nucleic Acids Res 47, D923-D929 (2019).21. Kluesner, M.G. et al. CRISPR-Cas9 cytidine and adenosine base editing of splice-sites mediates highly-efficient disruption of proteins in primary and immortalized cells. Nat Commun 12, 2437 (2021).15 22. Kim, E. & Hart, T. Improved analysis of CRISPR fitness screens and reduced off-target effects with the BAGEL2 gene essentiality classifier. Genome Med 13, 2 (2021).23. Pacini, C. et al. Integrated cross-study datasets of genetic dependencies in cancer. Nat Commun 12, 1661 (2021). 24. Nishimasu, H. et al. Engineered CRISPR-Cas9 nuclease with expanded targeting space. Science 20 361, 1259-1262 (2018). 25. Behan, F.M. et al. Prioritization of cancer therapeutic targets using CRISPR-Cas9 screens. Nature 568, 511-516 (2019). 26. Hornbeck, P.V. et al. PhosphoSitePlus, 2014: mutations, PTMs and recalibrations. Nucleic Acids Res 43, D512-20 (2015).25 27. Wagenaar, T.R. et al. Resistance to vemurafenib resulting from a novel mutation in the BRAFV600E kinase domain. Pigment Cell Melanoma Res 27, 124-33 (2014).28. Tian, J. et al. Combined PD-1, BRAF and MEK inhibition in BRAF(V600E) colorectal cancer: a phase 2 trial. Nat Med 29, 458-466 (2023).29. Flaherty, K.T. et al. Improved survival with MEK inhibition in BRAF-mutated melanoma. N Engl J30 Med 367, 107-14 (2012).30. Prahallad, A. et al. Unresponsiveness of colon cancer to BRAF(V600E) inhibition through feedback activation of EGFR. Nature 483, 100-3 (2012).31. Kopetz, S. et al. Encorafenib, Binimetinib, and Cetuximab in BRAF V600E-Mutated Colorectal Cancer. N Engl J Med 381, 1632-1643 (2019).35 32. Huijberts, S.C., van Geel, R.M., Bernards, R., Beijnen, J.H. & Steeghs, N. Encorafenib, binimetinib and cetuximab combined therapy for patients with BRAFV600E mutant metastatic colorectal cancer. Future Oncol 16, 161-173 (2020).33. Li, S. et al. Structural basis for inhibition of the epidermal growth factor receptor by cetuximab. Cancer Cell 7, 301-11 (2005).40 34. Khan, Z.M. et al. Structural basis for the action of the drug trametinib at KSR-bound MEK. Nature 588, 509-514 (2020). 35. Delaney, A.M., Printen, J.A., Chen, H., Fauman, E.B. & Dudley, D.T. Identification of a novel mitogen-activated protein kinase kinase activation domain recognized by the inhibitor PD 184352. Mol Cell Biol 22, 7593-602 (2002).45 36. Arena, S. et al. Emergence of Multiple EGFR Extracellular Mutations during Cetuximab Treatment in Colorectal Cancer. Clin Cancer Res 21, 2157-66 (2015).37. Emery, C.M. et al. MEK1 mutations confer resistance to MEK and B-RAF inhibition. Proc Natl Acad Sci U S A 106, 20411-6 (2009).38. Sale, M.J. et al. MEK1 / 2 inhibitor withdrawal reverses acquired resistance driven by BRAF(V600E)50 amplification whereas KRAS(G13D) amplification promotes EMT-chemoresistance. Nat Commun 10, 2030(2019). 39. Zhu, J., Woods, D., McMahon, M. & Bishop, J.M. Senescence of human fibroblasts induced by oncogenic Raf. Genes Dev 12, 2997-3007 (1998).40. Siravegna, G. et al. Clonal evolution and resistance to EGFR blockade in the blood of colorectal55 cancer patients. Nat Med 21, 795-801 (2015).41. Russo, M. et al. Tumor Heterogeneity and Lesion-Specific Response to Targeted Therapy in Colorectal Cancer. Cancer Discov 6, 147-153 (2016).42. Brammeld, J.S. et al. Genome-wide chemical mutagenesis screens allow unbiased saturation of the cancer genome and identification of drug resistance mutations. Genome Res 27, 613-625 (2017). 128 43. Sun, C. et al. Reversible and adaptive resistance to BRAF(V600E) inhibition in melanoma. Nature 508, 118-22 (2014). 44. Das Thakur, M. et al. Modelling vemurafenib resistance in melanoma reveals a strategy to forestall drug resistance. Nature 494, 251-5 (2013).5 45. Moriceau, G. et al. Tunable-combinatorial mechanisms of acquired resistance limit the efficacy of BRAF / MEK cotargeting but result in melanoma drug addiction. Cancer Cell 27, 240-56 (2015).46. Tate, J.G. et al. COSMIC: the Catalogue Of Somatic Mutations In Cancer. Nucleic Acids Res 47,D941-D947 (2019). 47. Schoffski, P. et al. A phase Ib study of pictilisib (GDC-0941) in combination with paclitaxel, with 10 and without bevacizumab or trastuzumab, and with letrozole in advanced breast cancer. Breast Cancer Res 20, 109 (2018).48. Skoulidis, F. et al. Sotorasib for Lung Cancers with KRAS p.G12C Mutation. N Engl J Med 384,2371-2381 (2021). 49. Janne, P.A. et al. Adagrasib in Non-Small-Cell Lung Cancer Harboring a KRAS(G12C) Mutation.15 N Engl J Med 387, 120-131 (2022).50. Awad, M.M. et al. Acquired Resistance to KRAS(G12C) Inhibition in Cancer. N Engl J Med 384,2382-2393 (2021). 51. Kim, D. et al. Pan-KRAS inhibitor disables oncogenic signalling and tumour growth. Nature (2023). 52. Pettitt, S.J. et al. Genome-wide and high-density CRISPR-Cas9 screens identify point mutations in20 PARP1 causing PARP inhibitor resistance. Nat Commun 9, 1849 (2018).53. Goodall, J. et al. Circulating Cell-Free DNA to Guide Prostate Cancer Treatment with PARP Inhibition. Cancer Discov 7, 1006-1017 (2017).54. Lin, K.K. et al. BRCA Reversion Mutations in Circulating Tumor DNA Predict Primary and Acquired Resistance to the PARP Inhibitor Rucaparib in High-Grade Ovarian Carcinoma. Cancer Discov 9, 210-21925 (2019). 55. Quigley, D. et al. Analysis of Circulating Cell-Free DNA Identifies Multiclonal Heterogeneity of BRCA2 Reversion Mutations Associated with Resistance to PARP Inhibitors. Cancer Discov 7, 999-1005(2017). 56. Tutt, A.N.J. et al. Adjuvant Olaparib for Patients with BRCA1- or BRCA2-Mutated Breast Cancer.30 N Engl J Med 384, 2394-2405 (2021).57. Gonzalez-Martin, A. et al. Niraparib in Patients with Newly Diagnosed Advanced Ovarian Cancer. NEngl J Med 381, 2391-2402 (2019).58. Garnett, M.J. et al. Systematic identification of genomic markers of drug sensitivity in cancer cells. Nature 483, 570-5 (2012).35 59. Gill, S.J. et al. Combinations of PARP Inhibitors with Temozolomide Drive PARP1 Trapping and Apoptosis in Ewing's Sarcoma. PLoS One 10, e0140988 (2015).60. Murai, J. et al. Trapping of PARP1 and PARP2 by Clinical PARP Inhibitors. Cancer Res 72, 5588-99 (2012). 61. Dawicki-McKenna, J.M. et al. PARP-1 Activation Requires Local Unfolding of an Autoinhibitory40 Domain. Mol Cell 60, 755-768 (2015).62. Ogden, T.E.H. et al. Dynamics of the HD regulatory subdomain of PARP-1; substrate access and allostery in PARP activation and inhibition. Nucleic Acids Res 49, 2266-2288 (2021).63. Ryan, K. et al. Dissecting the molecular determinants of clinical PARP1 inhibitor selectivity for tankyrase1. J Biol Chem 296, 100251 (2021).45 64. Vaclova, T. et al. Clinical impact of subclonal EGFR T790M mutations in advanced-stage EGFR- mutant non-small-cell lung cancers. Nat Commun 12, 1780 (2021).65. Cross, D.A. et al. AZD9291, an irreversible EGFR TKI, overcomes T790M-mediated resistance to EGFR inhibitors in lung cancer. Cancer Discov 4, 1046-61 (2014).66. Chmielecki, J. et al. Candidate mechanisms of acquired resistance to first-line osimertinib in EGFR-50 mutated advanced non-small cell lung cancer. Nat Commun 14, 1070 (2023).67. Chmielecki, J. et al. Analysis of acquired resistance mechanisms to osimertinib in patients with EGFR-mutated advanced non-small cell lung cancer from the AURA3 trial. Nat Commun 14, 1071 (2023).68. Thress, K.S. et al. Acquired EGFR C797S mutation mediates resistance to AZD9291 in non-small cell lung cancer harboring EGFR T790M. Nat Med 21, 560-2 (2015).55 69. Chen, P.J. et al. Enhanced prime editing systems by manipulating cellular determinants of editing outcomes. Cell 184, 5635-5652 e29 (2021).70. Ferreira da Silva, J. et al. Prime editing efficiency and fidelity are enhanced in the absence of mismatch repair. Nat Commun 13, 760 (2022).71. Mathis, N. et al. Predicting prime editing efficiency and product purity by deep learning. Nat 60 Biotechnol (2023). 129 72. Nelson, J.W. et al. Engineered pegRNAs improve prime editing efficiency. Nat Biotechnol 40, 402-410 (2022). 73. Chen, W.S. et al. Functional independence of the epidermal growth factor receptor from a domain required for ligand-induced internalization and calcium regulation. Cell 59, 33-43 (1989).5 74. Lovly, C. et al. Detection of Diverse EGFR c-terminal Truncations (C-trunc) and Sensitivity to Tyrosine Kinase Inhibitors (TKIs) in the Clinic. Journal of Thoracic Oncology 16, S604-S605 (2021).75. Dixit, A. et al. Perturb-Seq: Dissecting Molecular Circuits with Scalable Single-Cell RNA Profiling of Pooled Genetic Screens. Cell 167, 1853-1866 e17 (2016).76. Replogle, J.M. et al. Mapping information-rich genotype-phenotype landscapes with genome-scale10 Perturb-seq. Cell 185, 2559-2575 e28 (2022).77. Cooper, S.E. et al. scSNV-seq: high-throughput phenotyping of single nucleotide variants by coupled single-cell genotyping and transcriptomics. Genome Biol 25, 20 (2024).78. Zhu, S. et al. Guide RNAs with embedded barcodes boost CRISPR-pooled screens. Genome Biol 20, 20 (2019). 15 79. Watterson, A. & Coelho, M.A. Cancer immune evasion through KRAS and PD-L1 and potential therapeutic interventions. Cell Commun Signal 21, 45 (2023).80. Mugarza, E. et al. Therapeutic KRAS(G12C) inhibition drives effective interferon-mediated antitumor immunity in immunogenic lung cancers. Sci Adv 8, eabm8780 (2022).81. Coelho, M.A. et al. Oncogenic RAS Signaling Promotes Tumor Immunoresistance by Stabilizing20 PD-L1 mRNA. Immunity 47, 1083-1099 e6 (2017).82. Haas, L. et al. Acquired resistance to anti-MAPK targeted therapy confers an immune-evasive tumor microenvironment and cross-resistance to immunotherapy in melanoma. Nat Cancer 2, 693-708(2021). 83. Cattaneo, C.M. et al. Tumor organoid-T-cell coculture systems. Nat Protoc 15, 15-39 (2020).25 84. Dijkstra, K.K. et al. Generation of Tumor-Reactive T Cells by Co-culture of Peripheral Blood Lymphocytes and Tumor Organoids. Cell 174, 1586-1598 e12 (2018).85. El-Jawhari, J.J. et al. Blocking oncogenic RAS enhances tumour cell surface MHC class I expression but does not alter susceptibility to cytotoxic lymphocytes. Mol Immunol 58, 160-8 (2014).86. Ebert, P.J.R. et al. MAP Kinase Inhibition Promotes T Cell and Anti-tumor Activity in Combination30 with PD-L1 Checkpoint Blockade. Immunity 44, 609-621 (2016).87. Frederick, D.T. et al. BRAF inhibition is associated with enhanced melanoma antigen expression and a more favorable tumor microenvironment in patients with metastatic melanoma. Clin Cancer Res 19,1225-31 (2013). 88. Mimura, K. et al. The MAPK pathway is a predominant regulator of HLA-A expression in35 esophageal and gastric cancer. J Immunol 191, 6261-72 (2013).89. Sers, C. et al. Down-regulation of HLA Class I and NKG2D ligands through a concerted action of MAPK and DNA methyltransferases in colorectal cancer cells. Int J Cancer 125, 1626-39 (2009).90. Pallaseni, A. et al. Predicting base editing outcomes using position-specific sequence determinants. Nucleic Acids Res 50, 3551-3564 (2022).40 91. Arbab, M. et al. Determinants of Base Editing Outcomes from Target Library Analysis and Machine Learning. Cell 182, 463-480 e30 (2020).92. Algazi, A.P. et al. Continuous versus intermittent BRAF and MEK inhibition in patients with BRAF- mutated melanoma: a randomized phase 2 trial. Nat Med 26, 1564-1568 (2020).93. Sartore-Bianchi, A. et al. Circulating tumor DNA to guide rechallenge with panitumumab in45 metastatic colorectal cancer: the phase 2 CHRONOS trial. Nat Med 28, 1612-1618 (2022).94. Fujita, S., Masago, K., Katakami, N. & Yatabe, Y. Transformation to SCLC after Treatment with the ALK Inhibitor Alectinib. J Thorac Oncol 11, e67-72 (2016).95. Marcoux, N. et al. EGFR-Mutant Adenocarcinomas That Transform to Small-Cell Lung Cancer and Other Neuroendocrine Carcinomas: Clinical Outcomes. J Clin Oncol 37, 278-285 (2019).50 96. Geraghty, R.J. et al. Guidelines for the use of cell lines in biomedical research. Br J Cancer 111,1021-46 (2014). 97. Coelho, M.A. et al. BE-FLARE: a fluorescent reporter of base editing activity reveals editing characteristics of APOBEC3A and APOBEC3B. BMC Biol 16, 150 (2018).98. Fu, J. et al. Human cell based directed evolution of adenine base editors with improved efficiency.55 Nat Commun 12, 5897 (2021).99. Li, W. et al. MAGeCK enables robust identification of essential genes from genome-scale CRISPR / Cas9 knockout screens. Genome Biol 15, 554 (2014).100. Koeppel, J. et al. Prediction of prime editing insertion efficiencies using sequence features and DNA repair determinants. Nat Biotechnol (2023). 130 101. McCarthy, D.J., Campbell, K.R., Lun, A.T. & Wills, Q.F. Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics 33, 1179-1186 (2017).102. Michlits, G. et al. CRISPR-UMI: single-cell lineage tracing of pooled CRISPR-Cas9 screens. Nat Methods 14, 1191-1197 (2017).5 103. Squair, J.W. et al. Confronting false discoveries in single-cell differential expression. Nat Commun 12, 5692 (2021). 104. Schubert, M. et al. Perturbation-response genes reveal signaling footprints in cancer gene expression. Nat Commun 9, 20 (2018).105. Landais, Y. & Vallot, C. Multi-modal quantification of pathway activity with MAYA. Nat Commun 14,10 1668 (2023). 106. Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst 1, 417-425 (2015).107. Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat Biotechnol 36, 411-420 (2018).15 108. Scrucca, L., Fop, M., Murphy, T.B. & Raftery, A.E. mclust 5: Clustering, Classification and Density Estimation Using Gaussian Finite Mixture Models. R J 8, 289-317 (2016).109. Lun ATL, McCarthy DJ, Marioni JC (2016). “A step-by-step workflow for low-level analysis of single- cell RNA-seq data with Bioconductor.” F1000Res., 5, 2122. 110. Yoav Benjamini, Yosef Hochberg, Controlling the False Discovery Rate: A Practical and Powerful 20 Approach to Multiple Testing, Journal of the Royal Statistical Society: Series B (Methodological), Volume 57, Issue 1, January 1995, Pages 289–300. 111. Yoav Benjamini, Marina Bogomolov, Selective Inference on Multiple Families of Hypotheses, Journal of the Royal Statistical Society Series B: Statistical Methodology, Volume 76, Issue 1, January 2014, Pages 297–318. 25 112. Esposito, D. et al. MaveDB: an open-source platform to distribute and interpret data from multiplexed assays of variant effect. Genome Biol 20, 223 (2019).For standard molecular biology techniques, see Sambrook, J., Russel, D.W. Molecular Cloning, A Laboratory Manual.3 ed.2001, Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press 30

Claims

1. 131 Claims:

1. A method for predicting the therapeutic response of a cancer in a patient and / or for identifying a treatment for a patient who has been diagnosed as having or being likely to have cancer, the method comprising detecting the presence of one or more mitogen pathway protein variants in cancer cells of the 5 patient, wherein the mitogen pathway protein variants are selected from: a) variants that are associated with a C-terminal truncation of EGFR, and in particular a truncation after amino acids E1091 or L1038; b) EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; 10 c) KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; d) MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; e) MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; 15 f) BRAF variants selected from: V480A, K499E, K499R; g) PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A; and h) AKT1 variants selected from: E17K, G16E, wherein the presence of the one or more mitogen pathway protein variants are indicative of response to one or more of: MAP2K1 / 2 inhibitors, BRAF inhibitors, EGFR inhibitors, PI3K inhibitors, and KRAS 20 inhibitors; and optionally selecting the patient for treatment or recommending treatment with one or more therapies that the patient has been identified as likely to be responsive to or one or more therapies that are not therapies that the patient has been identified as likely to be resistant to. 25 2. The method of claim 1, wherein the variants are selected from: a) variants that are associated with a C-terminal truncation of EGFR after amino acids E1091 or L1038; b) KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; c) MAP2K1 variants selected from: I99T, L98P, S194P; 30 d) MAP2K2 variants selected from: Y134H; e) BRAF variants selected from: K499E, K499R.

3. The method of claim 1 or claim 2, wherein a patient who has been detected as having one or more mitogen pathway protein variants including KRAS E62 and / or E63 to K is predicted to be resistant to BRAF 35 inhibitors and / or EGFR inhibitors, optionally wherein: the patient is predicted to be resistant to the combination of a BRAF inhibitor and an EGFR inhibitor, optionally wherein the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab, and / or the method further comprises selecting the patient for treatment with a therapy that is not a BRAF inhibitor or EGFR inhibitor, optionally a MAP2K1 / 2 inhibitor, further optionally trametinib, and / or132 the patient who has been detected as having one or more mitogen pathway protein variants including KRAS E62 and / or E63 to K is predicted or detected as not having a mutation in the BRAF gene.

4. The method of any preceding claim, wherein a patient who has been detected as having one or 5 more mitogen pathway protein variants including KRAS K117 to E, R and / or D119 to G is predicted to be resistant to BRAF inhibitors and / or EGFR inhibitors, optionally wherein: the patient is predicted to be resistant to the combination of a BRAF inhibitor and an EGFR inhibitor, optionally wherein the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab or gefitinib, and / or 10 the method further comprises selecting the patient for treatment with a therapy that is not a BRAF inhibitor or EGFR inhibitor, optionally a MAP2K1 / 2 inhibitor, further optionally trametinib.

5. The method of any preceding claim, wherein a patient who has been detected as having one or more mitogen pathway protein variants including MAP2K2 Y134 to H is predicted to be resistant to one or 15 more of: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 inhibitors, and KRAS inhibitors, optionally wherein: the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab or gefitinib and / or the MAP2K1 / 2 inhibitor is trametinib and / or the KRAS inhibitor is a KRAS G12C inhibitor or a pan-KRAS inhibitor, optionally sotorasib or adagrasib;and / or the method further comprises selecting the patient for treatment with one or more of: BRAF inhibitors, 20 EGFR inhibitors, MAP2K1 / 2 inhibitors, and KRAS inhibitors using an intermittent treatment scheme, or selecting the patient for treatment with a therapy that includes a therapeutic that is not selected from: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 inhibitors, or KRAS inhibitors.

6. The method of any preceding claim, wherein a patient who has been detected as having one or 25 more mitogen pathway protein variants including MAP2K1 S194 to P is predicted to be resistant to one or more of: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 inhibitors, and KRAS inhibitors; and / or wherein a patient who has been detected as not having mitogen pathway protein variant MAP2K1 S194 to P and / or Y134H is selected for treatment with a combination of a MAP2K1 / 2 inhibitor and a T-cell therapy or vaccine; 30 optionally wherein: (i) the BRAF inhibitor is dabrafenib and / or the EGFR inhibitor is cetuximab or gefitinib and / or the MAP2K1 / 2 inhibitor is trametinib and / or the KRAS inhibitor is a KRAS G12C inhibitor or a pan-KRAS inhibitor, optionally sotorasib or adagrasib, and / or (ii) the method further comprising selecting the patient for treatment with one or more of: BRAF inhibitors, EGFR inhibitors, and / or MAP2K1 / 2 inhibitors, and / or KRAS inhibitors using an intermittent treatment scheme, or selecting the patient for treatment with 35 a therapy that is not limited to: BRAF inhibitors, EGFR inhibitors, MAP2K1 / 2 inhibitors, KRAS inhibitors, and / or a T-cell therapy or vaccine.

7. The method of any preceding claim, wherein a patient who has been detected as having one or more mitogen pathway protein variants including MAP2K1 L98 to P and / or I99 to T is predicted to be 40 resistant to one or more MAP2K1 / 2 inhibitors,133 optionally wherein the one or more MAP2K1 / 2 inhibitors is trametinib and / or the method further comprises selecting the patient for treatment with one or more MAP2K1 / 2 inhibitors using an intermittent therapeutic scheme, and / or selecting the patient for treatment with a drug that is not a MAP2K1 / 2 inhibitor. 5 8. The method of any preceding claim, wherein a patient who has been detected as having one or more mitogen pathway protein variants including a variant causing a C-terminal truncation of EGFR after amino acid E1091 or L1038 is predicted to be sensitive to one or more EGFR inhibitors, optionally wherein the one or more EGFR inhibitors are selected from: gefitinib, osimertinib, erlotinib, lapatinib, and cetuximab, and / or the method further comprises selecting the patient for treatment 10 with the one or more EGFR inhibitors.

9. The method of claim 8, wherein the variant causing a C-terminal truncation of EGFR after amino acid E1091 is a single nucleotide variant (SNV), optionally wherein the SNV is: i) a G to A mutation at location 7:55202626, or 15 ii) a T to C mutation at location 7:55202627, wherein the locations refer to locations in the GRCh38 reference genome assembly, and optionally wherein the SNV results in disruption of the splice donor site at the 5’ end of EGFR intron 27, and / or wherein the SNV results in creation of a premature stop codon. 20 10. The method of claim 8, wherein the variant causing a C-terminal truncation of EGFR after amino acid L1038 is a single nucleotide variant (SNV), optionally wherein the SNV is a G to A mutation at location 7:55201356, wherein the location refers to a location in the GRCh38 reference genome assembly, and optionally wherein the SNV results in disruption of the splice donor site at the 5’ end of EGFR intron 25, and / or wherein the SNV results in creation of a premature stop codon. 25 11. The method of any preceding claim, wherein a patient who has been detected as having one or more mitogen pathway protein variants including a variant BRAF K499 to E or R is predicted to be resistant to one or more of: BRAF inhibitors, EGFR inhibitors, and MAP2K1 / 2 inhibitors, optionally wherein the one or more EGFR inhibitors include gefitinib, the BRAF inhibitors include 30 cetuximab, and / or the MAP2K1 / 2 inhibitors include trametinib, optionally further comprising selecting the patient for treatment with a therapy that is not an EGFR inhibitor, BRAF inhibitor or MAP2K1 / 2 inhibitor.

12. The method of any preceding claim, further comprising treating the patient with the one or more 35 drugs that the patient has been selected or recommended for treatment with.

13. The method of any preceding claim, further comprising detecting the presence of the one or more mitogen pathway protein variants in a sample from said patient.134 14. A MAP2K1 / 2 inhibitor, optionally trametinib, for use in a method of treatment of cancer in a patient, the method comprising: a) identifying a tumour that is susceptible to treatment with said MAP2K1 / 2 inhibitor according to the method of claims 1-6 or 13, and 5 b) treating the patient whose tumour has been identified as susceptible to treatment with said MAP2K1 / 2 inhibitor in step (a) with said MAP2K1 / 2 inhibitor, optionally using an intermittent therapeutic scheme.

15. An EGFR inhibitor, optionally selected from the group of gefitinib and osimertinib, for use in a 10 method of treatment of cancer in a patient, the method comprising: a) identifying a tumour that is susceptible to treatment with said EGFR inhibitor according to the method of claims 1, 2, 5-10 or 13, and b) treating the patient whose tumour has been identified as susceptible to treatment with said EGFR inhibitor in step (a) with said EGFR inhibitor, optionally using an intermittent therapeutic 15 scheme.

16. A BRAF inhibitor, optionally dabrafenib, and / or an EGFR inhibitor, optionally cetuximab, for use in a method of treatment of cancer in a patient, the method comprising: a) identifying a tumour that is susceptible to treatment with said BRAF inhibitor and / or said EGFR 20 inhibitor according to the method of claims 1, 2, 5-7, or 13, and b) treating the patient whose tumour has been identified as susceptible to treatment with said BRAF inhibitor and / or said EGFR inhibitor in step (a) with said BRAF inhibitor and / or said EGFR inhibitor, optionally using an intermittent therapeutic scheme. 25 17. A KRAS inhibitor, optionally a KRAS G12C inhibitor or a pan-KRAS inhibitor, optionally sotorasib or adagrasib, for use in a method of treatment of cancer in a patient, the method comprising: a) identifying a tumour that is susceptible to treatment with said KRAS inhibitor according to the method of claims 1-7, 11 or 13, and b) treating the patient whose tumour has been identified as susceptible to treatment with said 30 KRAS inhibitor in step (a) with said KRAS inhibitor, optionally using an intermittent therapeutic scheme.

18. A PI3K inhibitor, optionally a pan-PI3K inhibitor, optionally pictilisib, for use in a method of treatment of cancer in a patient, the method comprising: 35 a) identifying a tumour that is susceptible to treatment with said PI3K inhibitor according to the method of claims 1-7, 11, or 13, and b) treating the patient whose tumour has been identified as susceptible to treatment with said PI3K inhibitor in step (a) with said PI3K inhibitor, optionally using an intermittent therapeutic scheme. 40135 19. The use of an anticancer agent in the preparation of a medicament for the treatment of a tumour having a mutation in a mitogen pathway protein, wherein a) the anticancer agent is a MAP2K1 / 2 inhibitor, optionally trametinib, and the mutation includes KRAS E62 and / or E63 to K, KRAS K117 to E, R and / or D119 to G, MAP2K2 Y134 to H, and / or 5 MAP2K1 S194 to P; b) the anticancer agent is an EGFR inhibitor, optionally selected from the group of gefitinib and osimertinib, and the mutation includes MAP2K2 Y134 to H, MAP2K1 S194 to P, MAP2K1 L98 to P and / or I99 to T, and / or a variant causing a C-terminal truncation of EGFR after amino acid E1091 or L1038; 10 c) the anticancer agent is a BRAF inhibitor, optionally dabrafenib, and / or an EGFR inhibitor, optionally cetuximab, and the mutation includes MAP2K2 Y134 to H, MAP2K1 S194 to P, MAP2K1 L98 to P and / or I99 to T; d) the anticancer agent is a KRAS inhibitor, optionally a KRAS G12C inhibitor or a pan-KRAS inhibitor, optionally sotorasib or adagrasib, and the mutation includes KRAS E62 and / or E63 to K, KRAS 15 K117 to E, R and / or D119 to G, MAP2K2 Y134 to H, and / or MAP2K1 S194 to P, MAP2K1 L98 to P and / or I99 to T, and / or BRAF K499 to E or R; e) the anticancer agent is a PI3K inhibitor, optionally a pan-PI3K inhibitor, optionally pictilisib, and the mutation includes KRAS E62 and / or E63 to K, KRAS K117 to E, R and / or D119 to G, MAP2K2 Y134 to H, and / or MAP2K1 S194 to P, MAP2K1 L98 to P and / or I99 to T, and / or BRAF K499 to E 20 or R.

20. A nucleic acid probe capable of specifically hybridising to nucleic acid encoding a mutated mitogen pathway protein or fragment thereof incorporating one or more mitogen pathway protein variants, wherein the variants are selected from: 25 a) variants that are associated with a C-terminal truncation of EGFR, and in particular a truncation after amino acids E1091 or L1038; b) EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; c) KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; 30 d) MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, D208N, V211A; e) MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; f) BRAF variants selected from: V480A, K499E, K499R; 35 g) PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A; and h) AKT1 variants selected from: E17K, G16E.

21. A method of identifying a mitogen pathway protein mutation in a sample comprising contacting nucleic acid from said sample with a nucleic acid probe according to claim 20, and detecting said 40 hybridisation.136 22. A method of screening for compounds or compositions that overcome resistance of a cell that incorporates a mitogen pathway protein variant that is associated with response to one or more drugs, wherein the mitogen pathway protein variants are selected from: 5 a) EGFR variants selected from: D46N, F436S, F436L, S437P, FS436-437PP, S464L, N493D, N493S, R494G, I491M, S720F, T790M, D1006N, E1005K, E1004K, D1012N, D1014N, E1015K; b) KRAS variants selected from: E62K, E63K, K117E, K117R, D119G; c) MAP2K1 variants selected from: Q45R, Q46R, F53L, F53S, K57E, K57R, Q58R, I99T, L98P, I111T, Q110H, I112T, L115T, H119P, E120G, N122S, N122D, Y130C, Y130H, S194P, E203K, 10 D208N, V211A; d) MAP2K2 variants selected from: Q60R, K61R, Y134H, V195A, S198P, I208T, L210P; e) BRAF variants selected from: V480A, K499E, K499R; f) PIK3CA variants selected from: E545K, E542K, E547K, E970G, Q969R, T972A; and g) AKT1 variants selected from: E17K, G16E, 15 wherein the presence of the one or more mitogen pathway protein variants are indicative of resistance to one or more of: MAP2K1 / 2 inhibitors, BRAF inhibitors, EGFR inhibitors, PI3K inhibitors, and KRAS inhibitors; and wherein the method comprises: contacting a cell that expresses said one or more mitogen pathway protein variants with one or more candidate compounds or compositions and comparing the effect of the one or more candidate 20 compounds or compositions on said cells with one or more control conditions, and / or contacting said one or more mitogen pathway protein variants with one or more candidate compounds or compositions and comparing the effect of the one or more candidate compounds or compositions on said mitogen pathway protein variant(s) with one or more control conditions, and determining the effect of said one or more candidate compounds or compositions on activity of the 25 mitogen pathway protein variants.

23. The method of claim 22, wherein the one or more candidate compounds or compositions comprise an antigen-binding molecule, optionally wherein the antigen-binding molecule is capable of binding to a linear or conformational epitope of EGFR, MAP2K1 / 2, BRAF, PI3KCA, AKT1 or KRAS, optionally wherein 30 the epitope comprises one or more of said mitogen pathway protein variant positions.

24. The method of claim 22 or claim 23, wherein determining the effect of said one or more candidate compounds or compositions comprises detecting a change in the proliferation rate of cells that expresses said one or more mitogen pathway protein variants in the presence of a candidate compound or 35 composition, whereby a reduction in proliferation rate of said cells compared to a control, or compared to the proliferation rate in the absence of said candidate compound or composition, indicates said candidate compound is an inhibitor of mutant mitogen pathway protein signalling, optionally further comprising selecting a subset of candidate compounds which cause a reduction in said proliferation rate. 40137 25. The method of any one of claims 22-24, wherein determining the effect of said one or more candidate compounds or compositions comprises determining the extent of phosphorylation of downstream mitogen pathway proteins, optionally via Western blotting, optionally further comprising selecting a subset of candidate compounds or compositions which 5 cause a reduction in the extent of phosphorylation of downstream mitogen pathway proteins.

Images