Biomarkers for diagnosing non-small cell lung cancer (NSCLC)
The analysis of peptide structure data using a machine-learning model addresses the challenge of late NSCLC detection by enabling early and accurate diagnosis through the classification of NSCLC samples, improving treatment timing and outcomes.
Patent Information
- Application Number
- PCT/US2023/075858
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-25
- Filing Date
- 2023-10-03
- Publication Date
- 2024-11-14
AI Technical Summary
Current methods for diagnosing non-small cell lung cancer (NSCLC) are inadequate due to late detection and limited effective therapeutic options, with early-stage symptoms often missed or mistaken for other conditions, leading to substantial drops in survival rates once the cancer metastasizes.
A method involving the analysis of peptide structure data from biological samples using a machine-learning model, specifically a regularized regression model like LASSO, to classify samples as having NSCLC or a healthy state, utilizing peptide structures identified through liquid chromatography-mass spectrometry (LC-MS) and multiple reaction monitoring mass spectrometry (MRM-MS), allowing for early and unambiguous diagnosis.
This approach enables early and accurate diagnosis of NSCLC, potentially improving treatment outcomes by allowing for timely intervention before metastasis, using peptide structures and their quantification data to determine disease indicators and classify samples with high predictive performance.
Smart Images

Figure US2023075858_14112024_PF_FP_ABST
Abstract
Description
BIOMARKERS FOR DIAGNOSING NON-SMALL CELL LUNG CANCER(NSCLC)CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims priority from U.S. Provisional Application No. 63 / 501,145, filed May 9, 2023, and U.S. Provisional Application No. 63 / 504,347, filed May 25, 2023, the contents of which are incorporated by reference in their entirety.SUBMISSION OF SEQUENCE LISTING ON ASCII TEXT FILE
[0002] The specification further incorporates by reference the Sequence Listing submitted herewith. The Sequence Listing .xml file, identified as VENN-00061 WO, is 311,716 bytes in size and was created on September 29, 2023. The Sequence Listing, electronically filed herewith, does not extend beyond the scope of the specification, and does not contain new matter.FIELD OF THE INVENTION
[0003] The present disclosure generally relates to methods and systems for diagnosing and / or treating non-small cell lung cancer (NSCLC). More particularly, the present disclosure relates to analyzing quantification data for a set of peptide structures detected in a biological sample obtained from a subject for use in a diagnostic assessment of the subject’s disease state (e.g., healthy state or NSCLC state) relating to a disease progression and / or treating the subject.BACKGROUND OF THE INVENTION
[0004] Lung cancer is the most common cause of cancer death worldwide, with an estimated 1.6 million deaths each year. For lung cancer cases in the U.S., more than 80% are attributable to tobacco smoking. Despite being a preventable disease, lung cancer remains one of the most common and lethal cancers globally due to treatment challenges and limited effective therapeutic options.
[0005] Most lung cancer patients (about 85%) are diagnosed with non-small cell lung cancer (NSCLC). Diagnosis is usually confirmed with various imaging techniques (e.g., chestx-ray, CT scan) and fluid analyses (e.g., sputum cytology, thoracentesis). If lung cancer is suspected, a biopsy is collected for morphological and molecular sample analysis. Additional diagnostic options are available including an endoscopic ultrasound, bronchoscopy, blood testing, and immunohistochemistry, each having contextual advantages. Early detection and diagnosis of NSCLC are key for effective treatment, yet early-stage symptoms are often missed or mistaken for other illness (e.g., pneumonia or a partially collapsed lung). Consequently, the patient may only notice symptoms when the cancer has metastasized regionally in the lungs or across the body, wherein the 5-year survival rate drops substantially, to less than 10%.
[0006] In view of the current treatment options, NSCLC is not considered a curable disease once it progresses to a stage III (e.g., larger tumor, potential for metastasis). Common treatments for NSCLC include surgery, for healthy individuals, wherein the cancer is physically removed by removing part of the lung (e.g., lobectomy) or removing the whole lung (e.g., pneumonectomy). Platinum-based therapies are common chemotherapy options, wherein platinum-coordination compounds, such as cisplatin (CDDP) and carboplatin (CBDCA are administered in combination with other cancer therapeutics. More recently, immunotherapies and patient- specific therapies have been used to improve treatment for specific NSCLC patient subgroups, but platinum-based therapies remain a therapeutic cornerstone for NSCLC treatment.
[0007] These NSCLC therapies can slow cancer progression, but early diagnosis is key for patient treatment and overall survival. Given the global burden and life threatening consequences of delayed NSCLC detection, early and unambiguous diagnosis and effective treatment strategies are imperative. Under certain circumstances, the diagnosis of NSCLC using a blood sample may be easier to do than performing a diagnostic test that uses a biopsy sample obtained from a lung.BRIEF SUMMARY OF THE INVENTION
[0008] In some embodiments, provided herein is a method of classifying a biological sample obtained from a subject with respect to a plurality of states associated with non-small cell lung cancer (NSCLC). In some embodiments, the method includes receiving peptide structure data corresponding to a set of proteins in the biological sample, inputtingquantification data identified from the peptide structure data for a set of peptide structures into a machine-learning model trained to identify a disease indicator based on the quantification data. In some embodiments, the set of peptide structures includes at least one peptide structure identified from a plurality of peptide structures in Table 3. In some embodiments, the method further includes identifying, by the machine-learning model, the disease indicator; and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the identified disease indicator. In some embodiments, the at least one peptide structure comprises a glycopeptide. In some embodiments, the set of proteins comprises one or more glycoproteins.
[0009] Also provided herein is a method of detecting the presence of non-small cell lung cancer (NSCLC) in a subject. In some embodiments, the method includes receiving peptide structure data corresponding to a set of proteins in a biological sample obtained from a subject. In some embodiments, the peptide structure data includes at least one peptide structure from Table 3. In some embodiments, the method further includes inputting quantification data identified from the peptide structure data for a set of peptide structures into a machine-learning model trained to identify a disease indicator based on the quantification data, and detecting the presence of NSCLC in response to a determination that the identified disease indicator falls within a selected range associated with NSCLC. In some embodiments, the at least one peptide structure comprises a glycopeptide. In some embodiments, the set of proteins comprises one or more glycoproteins.
[0010] In some embodiments, the plurality of states includes at least one of an NSCLC state or a healthy state. In some embodiments, the machine-learning model includes a regularized regression model. In some embodiments, the regularized regression model includes a least absolute shrinkage and selection operator (LASSO) regression model.
[0011] In some embodiments, the quantification data for a peptide structure of the set of peptide structures includes at least one of an abundance, a relative abundance, a normalized abundance, or a differential abundance. In some embodiments, the quantification data for a peptide structure of the set of peptide structures includes at least one of a relative quantity, an adjusted quantity, a normalized quantity, a relative concentration, an adjusted concentration, or a normalized concentration.
[0012] In some embodiments, the quantification data is generated using a liquid chromatography-mass spectrometry (LC-MS) system. In some embodiments, the peptide structure data is generated using multiple reaction monitoring mass spectrometry (MRM- MS). In some embodiments, the machine-learning model was trained utilizing a portion of the quantification data corresponding to a set of peptide structures that is a subset of the panel of peptide structures to determine which state of the plurality of states the biological sample from the subject corresponds. In some embodiments, the biological sample comprises at least one of blood, serum, or plasma.
[0013] In some embodiments, the methods further include performing a differential expression analysis using the quantification data for the plurality of subjects.
[0014] Also provided herein is a method of treating non- small cell lung cancer (NSCLC) in a subject. In some embodiments, method includes receiving peptide structure data corresponding to a set of proteins in the biological sample obtained from a subject. In some embodiments, the peptide structure data comprises at least one peptide structure from Table 3. In some embodiments, the method further includes inputting quantification data for the at least one peptide structure into a machine-learning model trained to generate disease indicator for NSCLC based on the quantification data, identifying, by the machine-learning model, the disease indicator, and determining at least one of a plurality of treatment regimens to treat NSCLC based upon the disease indicator. In some embodiments, the method further includes administering a selected treatment regimen to the subject. In some embodiments, the set of proteins comprises one or more glycoproteins.
[0015] Also provided herein is a method of treating non- small cell lung cancer (NSCLC) in a subject. In some embodiments, method includes receiving peptide structure data corresponding to a set of proteins in the biological sample, inputting quantification data identified from the peptide structure data for a set of peptide structures into a machinelearning model trained to identify a disease indicator based on the quantification data. In some embodiments, the peptide structure data includes at least one peptide structure identified from a plurality of peptide structures in Table 3. In some embodiments, the method further includes identifying, by the machine-learning model, the disease indicator, determining a classification for NSCLC based upon the identified disease indicator, and determining at least one of a plurality of treatment regimens to treat NSCLC based upon theclassification. In some embodiments, the method further includes administering a selected treatment regimen to the subject. In some embodiments, the set of proteins comprises one or more glycoproteins.
[0016] Also provided herein is a method of diagnosing an individual with non-small cell lung cancer (NSCLC). In some embodiments, the method includes detecting the presence or amount of at least one peptide structure structures from Table 3 or Table 8, inputting a quantification of the detected at least one peptide structure into a machine-learning model trained to generate a class label, determining if the class label is above or below a threshold for a classification, identifying a diagnostic classification for the individual based on whether the class label is above or below a threshold for the classification, and diagnosing the individual as having NSCLC based on the diagnostic classification.
[0017] In some embodiments, the quantification data is generated using a liquid chromatography-mass spectrometry (LC-MS) system. In some embodiments, the peptide structure data is generated using multiple reaction monitoring mass spectrometry (MRM- MS). In some embodiments, the amount of at least one peptide structure is none, or below a detection limit. In some embodiments, the NSCLC is one of early-stage or late-stage NSCLC. In some embodiments, the NSCLC is one of stage I NSCLC, stage II NSCLC, stage III NSCLC, or stage IV NSCLC. In some embodiments, the at least one peptide structure comprises three or more peptide structures identified in Table 3 or Table 8. In some embodiments, the at least one peptide structure comprises at least one peptide comprising the sequence set forth in any one of SEQ ID NOs: 27-99. In some embodiments, the individual is determined have a healthy state, wherein a healthy state comprises the absence of NSCLC.
[0018] In some embodiments, the methods further include assessing one or more risk factor or clinical indicators of NSCLC. In some embodiments, the methods further include generating a report that includes a diagnosis based on the corresponding state detected for the subject.
[0019] Also provided herein is a method of training a model to diagnose a subject with one of a plurality of states associated with non-small cell lung cancer (NSCLC). In some embodiments, the method includes receiving quantification data for a panel of peptide structures for a plurality of subjects diagnosed with the plurality of states associated withNSCLC, and training a machine-learning model to determine a state of the plurality of states a biological sample from the subject based on the quantification data.
[0020] In some embodiments, training the machine-learning model to determine the state of the plurality of states further includes training the machine-learning model to generate a class label for the state of the plurality of states. In some embodiments, training the machinelearning model to determine the state of the plurality of states comprises training and evaluating the machine-learning model based on one or more of: a first set of peptide structure coefficient from Table 7 for all stages of NSCLC; a second set of peptide structure coefficients from Table 7 for early-stage NSCLC; and a third set of peptide structure coefficients from Table 7 for late-stage NSCLC. In some embodiments, the first set of peptide structure coefficients comprise the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58. In some embodiments, the second set of peptide structure coefficients comprise the amino acid sequence of SEQ ID NOs: 27, 30, 36, 41, 43-45, 47, 50- Sl, 53-58. In some embodiments, the third set of peptide structure coefficients comprise the amino acid sequence of SEQ ID NOs: 28-29, 31-33, 36-40, 42, 44-47, 49, 57-60.
[0021] In some embodiments, the plurality of states comprises at least one of a NSCLC state or a healthy state. In some embodiments, the machine-learning model comprises a regularized regression model. In some embodiments, the regularized regression model comprises a least absolute shrinkage and selection operator (LASSO) regression model.
[0022] In some embodiments, at least one of the peptide structures comprises a glycopeptide.
[0023] In some embodiments, the at least one peptide structure comprises at least one, at least two, at least three, at least five, at least 10, at least 15, at least 20, or at least 25 different peptides comprising the sequence set forth in any one of SEQ ID NOs: 27-60.
[0024] In some embodiments, the at least one peptide structure comprises at least one, at least two, at least three, at least five, at least 10, at least 15, at least 20, or at least 25 different peptides comprising the sequence set forth in any one of SEQ ID NOs: 61-99.
[0025] In some embodiments, the at least one peptide structure comprises at least one, at least two, at least three, at least five, at least 10, or at least 15 different peptides comprising the sequence set forth in any one of SEQ ID NOs: 32, 34-37, 42, 44, 47-48, 50-58.
[0026] In some embodiments, the at least one peptide structure comprises at least one, at least two, at least three, at least five, at least 10, at least 15, at least 20, or at least 25 different peptides comprising the sequence set forth in any one of SEQ ID NOs: 27-60. In some embodiments, the at least one peptide structure comprises at least one, at least two, at least three, at least five, at least 10, at least 15, at least 20, or at least 25 different peptides comprising the sequence set forth in any one of SEQ ID NOs: 61-99. In some embodiments, the at least one peptide structure comprises at least one, at least two, at least three, at least five, at least 10, or at least 15 different peptides comprising the sequence set forth in any one of SEQ ID NOs: 32, 34-37, 42, 44, 47-48, 50-58.
[0027] In some embodiments, the at least one peptide structure comprises a peptide sequence and a glycan structure, wherein the glycan structure is attached to a linking site position in the peptide sequence in accordance with Table 3. In some embodiments, the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 3, wherein the glycan structure comprises a symbol structure in accordance with the glycan structure GL number according to Table 3, Table 4 A, and Table 4B. In some embodiments, the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 3, wherein the glycan structure comprises a composition in accordance with the glycan structure GL number, Table 3, Table 4A, and Table 4B. In some embodiments, a rightmost N-acetylgalactosamine (open square) of the glycan structure in Table 4A is attached to a linking site position in the peptide sequence in accordance with Table 3. In some embodiments, a bottommost N- acetylglucosamine (dark square) of the glycan structure in Table 4B is attached to a linking site position in the peptide sequence in accordance with Table 3.
[0028] In some embodiments, provided herein is a composition comprising one or more peptide structures from Table 3. In some embodiments, the at least one peptide structure comprises a peptide sequence and a glycan structure, wherein the glycan structure is attached to a linking site position in the peptide sequence in accordance with Table 3. In some embodiments, the glycan structure of the peptide sequence corresponds to a glycan structureGL number in accordance with Table 3, wherein the glycan structure comprises a symbol structure in accordance with the glycan structure GL number according to Table 3, Table 4 A, and Table 4B. In some embodiments, the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 3, wherein the glycan structure comprises a composition in accordance with the glycan structure GL number, Table 3, Table 4A, and Table 4B. In some embodiments, a rightmost N-acetylgalactosamine (GalNAc) of the glycan structure in Table 4 A is attached to a linking site position in the peptide sequence in accordance with Table 3. In some embodiments, a bottommost N- acetylglucosamine (GlcNAc) of the glycan structure in Table 4B is attached to a linking site position in the peptide sequence in accordance with Table 3.
[0029] In some embodiments, provided herein is a composition comprising one or more peptides comprising the sequence set forth in SEQ ID NOs: 27-99. In some embodiments, the one or more peptides comprise one or more glycopeptides.BRIEF DESCRIPTION OF THE DRAWINGS
[0030] FIG. 1 shows an exemplary workflow for the detection of peptide structures associated with a disease state for use in diagnosis and / or treatment in accordance with one or more embodiments.
[0031] FIG. 2A shows a schematic diagram of a preparation workflow in accordance with one or more embodiments.
[0032] FIG. 2B shows a process of data acquisition in accordance with one or more embodiments.
[0033] FIG. 3 is a block diagram of an analysis system in accordance with one or more embodiments.
[0034] FIG. 4 is a block diagram of a computer system in accordance with one or more embodiments.
[0035] FIG. 5 is a flowchart of a process for classifying a biological sample obtained from a subject with respect to a plurality of states associated with NSCLC in accordance with one or more embodiments.
[0036] FIG. 6 is a flowchart of a process for detecting the presence of one of a plurality of states associated with NSCLC in a subject in accordance with one or more embodiments.
[0037] FIG. 7 is a flowchart of a process for determining one or more of a plurality of treatment regimens for treating NSCLC in a subject in accordance with one or more embodiments.
[0038] FIG. 8 is a flowchart of a process for training a model to diagnose a subject with one of a plurality of states associated with NSCLC in accordance with one or more embodiments.
[0039] FIG. 9 is a flowchart of a process for treating NSCLC in a subject in accordance with one or more embodiments.
[0040] FIG. 10 is a flowchart of a process for diagnosing an individual with NSCLC in accordance with one or more embodiments.
[0041] FIG. 11 shows an experimental workflow for sample preparation and analysis for either plasma or serum.
[0042] FIG. 12 is a model performance plot for healthy and all NSCLC stages (1-4) samples illustrating the predictive performance of a regularized regression model (e.g., LASSO regression model) on the training data sets and test data sets.
[0043] FIG. 13 illustrates a receiver-operating-characteristic (ROC) curve / area under curve (AUC) for the regularized regression model (e.g., LASSO regression model) for a healthy control and all NSCLC stages (1-4) samples.
[0044] FIG. 14 is a model performance plot for healthy and early-stage NSCLC (stages 1-2) samples illustrating the predictive performance of regularized regression model (e.g., LASSO regression model) on the training data sets and test data sets.
[0045] FIG. 15 illustrates a receiver-operating-characteristic (ROC) curve / area under curve (AUC) for the regularized regression model (e.g., LASSO regression model) for healthy and early-stage NSCLC (stages 1-2) samples.
[0046] FIG. 16 is a model performance plot for healthy and late-stage NSCLC (stages 3- 4) samples illustrating the predictive performance of regularized regression model (e.g., LASSO regression model) on the training data sets and test data sets.
[0047] FIG. 17 illustrates a receiver-operating-characteristic (ROC) curve / area under curve (AUC) for the regularized regression model (e.g., LASSO regression model) for healthy and late-stage NSCLC (stages 3-4) samples.
[0048] FIG. 18 illustrates a targeted LC-MS-based glycoproteomic quantification workflow for sample processing implemented for this study.
[0049] FIG. 19A is a scatterplot of the top two principal components generated via PC A of the full assay by phenotype groups.
[0050] FIG. 19B shows intersection sizes of biomarkers achieving (black dots) or not achieving (grey dots) statistical significance at FDR<0.05 in three differential expression contrasts.
[0051] FIG. 20 is a heatmap showing the 50 biomarkers that are significantly differentially expressed at FDR<0.05 between healthy controls and NSCLC samples, COPD and NSCLC samples, but not between healthy controls and COPD samples in the full sample set.
[0052] FIG. 21 A shows receiver operating characteristic (ROC) curves representing discriminatory performance for all possible thresholds of the developed binary classifier, stratified by sample set.
[0053] FIG. 2 IB shows distributions of predicted probabilities in each phenotype and NSCLC stage group, stratified by sample set.
[0054] FIG. 22 shows a heatmap of biomarker expression for all patients in the held-out test set, grouped by true phenotype, and biomarkers filtered to only those that were retained in the primary multivariate classifier.
[0055] Figure 23 shows distributions of predicted probabilities of a secondary multivariate classifier trained only on healthy controls and NSCLC samples, stratified by phenotype and sample set. This classifier did not train on any COPD samples.
[0056] FIG. 24A is a chart showing a number of glycoforms represented in the final classifier by protein name.
[0057] FIG. 24B is a pie chart breakdown by category of 37 glycoforms of 42 markers retained in the final classifier.DETAILED DESCRIPTION OF THE INVENTION
[0058] Provided herein are methods useful for diagnosing NSCLC based upon one or more biomarkers. In some embodiments, the diagnosis is based upon the presence, absence, and / or amount of one or more peptide structures comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, a machine-learning model is used to classify the sample with respect to a state associated with NSCLC, such as NSCLC or a healthy state.I. Definitions
[0059] As used herein, the term “plurality” is more than 1 and may be 2, 3, 4, 5, 6, 7, 8, 9, 10, or more.
[0060] As used herein, the term “set of’ means one or more. For example, a set of items includes one or more items.
[0061] As used herein, the phrase “at least one of,” when used with a list of items, means different combinations of one or more of the listed items may be used and only one of the items in the list is required to be included. The item may be a particular object, thing, step, operation, process, or category. In other words, “at least one of’ means any combination of items or number of items may be used from the list, but not all of the items in the list may be required. For example, without limitation, “at least one of item A, item B, and item C” intends and includes any of item A; item A and item B; item B; item A, item B, and item C; item B and item C; item C; and item A and C. It is understood that “at least one of’ includes instance where more than one of any listed item is present. For example, and without limitation, at least one of item A, item B, and item C include an embodiment in which two of item A is present, one of item B is present, and ten of item C is present.
[0062] As used herein, “substantially” means sufficient to work for the intended purpose. The term “substantially” thus allows for minor, insignificant variations from an absolute or perfect state, dimension, measurement, result, or the like such as would be expected by a person of ordinary skill in the field but that do not appreciably affect overall performance.
[0063] The term “amino acid,” as used herein, generally refers to any organic compound that includes an amino group (e.g., -NH2), a carboxyl group (-COOH), and a side chain group (R) which varies based on a specific amino acid. Thus, “amino acid” includes organic compounds of the formula NH2-CH(R)-COOH where R represents an amino acid side chain group. In some instance R represents the side chain of a natural amino acid. Amino acids can be linked using peptide bonds.
[0064] The term “alkylation,” as used herein, generally refers to the transfer of an alkyl group from one molecule to another. In various embodiments, alkylation is used to react with reduced cysteines to prevent the re-formation of disulfide bonds after reduction has been performed.
[0065] The term “linking site” or “glycosylation site” as used herein generally refers to the location where a sugar molecule of a glycan or glycan structure is directly bound (e.g., covalently bound) to an amino acid of a peptide, a polypeptide, or a protein. For example, the linking site may be an amino acid residue and a glycan structure may be linked via an atom of the amino acid residue. Non-limiting examples of types of glycosylation can include N-linked glycosylation, O-linked glycosylation, C-linked glycosylation, S-linked glycosylation, and glycation. N-linked glycosylation can include a glycan attached to an asparagine. O-linked glycosylation can include a glycan attached to either a serine or a threonine.
[0066] The term “biomarker,” as used herein, generally refers to any measurable substance taken as a sample from a subject whose presence, absence and / or amount is indicative of some phenomenon. Non-limiting examples of such phenomenon can include a disease state, a condition, or exposure to a compound or environmental condition. In various embodiments described herein, biomarkers may be used for diagnostic purposes (e.g., to diagnose a disease state, a health state, an asymptomatic state, a symptomatic state, etc.). Theterm “biomarker” may be used interchangeably with the term “marker.” Biomarkers include peptide structures such as those listed in Table 3 and Table 8.
[0067] The term “denaturation,” as used herein, generally refers to protein unfolding. Non-limiting examples include proteins or nucleic acids being exposed to an external compound or environmental condition such as acid, base, temperature, pressure, radiation, etc.
[0068] The term “denatured protein,” as used herein, generally refers to a protein that loses quaternary structure, tertiary structure, and secondary structure which is present in its native state.
[0069] The terms “digestion” or “enzymatic digestion,” as used herein, generally refer to breaking apart a polymer (e.g., cutting a polypeptide at a cut site). Proteins may be digested in preparation for mass spectrometry using trypsin digestion protocols. Proteins may be digested using other proteases in preparation for mass spectrometry if access is limited to cleavage sites.
[0070] The term “disease progression,” as used herein, refers to a progression of a disease from no disease or a less advanced form of disease to a more advanced (e.g., severe) form of the disease. A disease progression may include any number of stages of the disease.
[0071] The term “disease state” as used herein, generally refers to a condition that affects the structure or function of an organism. Disease states can include, for example, stages of a disease progression. Disease states can include any state of a disease whether symptomatic or asymptomatic. Disease states can cause minor, moderate, or severe disruptions in the structure or function of a subject. Disease state includes NSCLC, early-stage NSCLC, latestage NSCLC, severe NSCLC, disposition or likelihood of NSCLC, or normal or healthy state with respect to NSCLC.
[0072] The terms “glycan” or “polysaccharide” as used herein, both generally refer to a carbohydrate residue of a glycoconjugate, such as the carbohydrate portion of a glycopeptide, glycoprotein, glycolipid, or proteoglycan. Glycans can include monosaccharides.
[0073] The term “glycoprotein” or “glycopolypeptide” as used herein, generally refers to a protein having at least one glycan residue bonded thereto. In some examples, aglycoprotein is a protein with at least one oligosaccharide chain covalently bonded thereto.Examples of glycoproteins, include but are not limited SEQ ID NOs: 1-7, 9-26 and 100.
[0074] The term “glycopeptide” as used herein, refers to a fragment of a glycoprotein, unless specified otherwise to the contrary. In various embodiments, glycopeptides comprise carbohydrate moieties (e.g., one or more glycans) covalently attached to a side chain (i.e., R group) of an amino acid residue. Examples of glycopeptides, include but are not limited to the glycopeptides provided in Table 3 and Table 8. Examples of glycopeptides, include but are not limited to SEQ ID NOs: 27-99 (that are not denoted as non-glycosylated).
[0075] The term “liquid chromatography,” as used herein, generally refers to a technique used to separate a sample into parts. Liquid chromatography can be used to separate, identify, and quantify components.
[0076] The term “mass spectrometry” as used herein, generally refers to an analytical technique used to identify molecules by measuring a mass-to-charge (m / z) ratio. In various embodiments described herein, mass spectrometry can be involved in characterization and sequencing of proteins as well as to determine the presence, absence and / or abundance or peptides or proteins.
[0077] The term “m / z” or “mass-to-charge ratio” as used herein, generally refers to an output value from a mass spectrometry instrument. In various embodiments, m / z can represent a relationship between the mass of a given ion and the number of elementary charges that it carries. The “m” in m / z stands for mass and the “z” stands for charge. In some embodiments, m / z can be displayed on an x-axis of a mass spectrum.
[0078] The term “peptide,” as used herein, refers to amino acids linked by peptide bonds less than 50 amino acids in length. Peptides can include amino acid chains shorter than 10 residues, including, oligopeptides, dipeptides, tripeptides, and tetrapeptides. Peptides includes peptides comprising consisting of, or consisting essentially of the peptide structures provided in Table 3 and Table 8.
[0079] The terms “protein” or “polypeptide” or may be used interchangeably herein and refer to a polymer in which the monomers are amino acid residues that are joined together through amide bonds of at least 50 amino acid residues in length. Proteins may be digestedin preparation for mass spectrometry using trypsin digestion protocols. Proteins may be digested using other proteases in preparation for mass spectrometry if access is limited to cleavage sites. Proteins can include glycoproteins, which are proteins that contain at least one glycan residue bonded thereto.
[0080] The term “peptide structure,” as used herein, generally refers to peptides or a portion thereof or glycopeptides or a portion thereof. In various embodiments described herein, a peptide structure can include any molecule comprising at least two amino acids in sequence. A peptide structure of a glycopeptide includes description of the peptide amino acids sequence as well as the location and identity of the associated glycan.
[0081] The term “reduction,” as used herein, generally refers to the gain of an electron by a substance. In various embodiments, reduction may be used to break disulfide bonds between two cysteines.
[0082] The term “sample” and “biological sample” as used herein, generally refers to a sample obtained from a subject of interest. The sample may include a cell sample. The sample may include a cell line or cell culture sample. The sample can include one or more cells. The sample can include one or more microbes. The sample may include a nucleic acid sample or protein sample. The sample may also include a carbohydrate sample or a lipid sample. The sample may be derived from another sample. The sample may include a tissue sample, such as a biopsy, core biopsy, needle aspirate, or fine needle aspirate. The sample may include a fluid sample, such as a blood sample, urine sample, or saliva sample. The sample may include a skin sample. The sample may include a cheek swab. The sample may include a plasma or serum sample. The sample may include a cell free sample. A cell-free sample may include extracellular polynucleotides. The sample may originate from blood, plasma, serum, urine, saliva, mucosal excretions, sputum, stool, or tears. The sample may originate from red blood cells or white blood cells. The sample may originate from feces, spinal fluid, CNS fluid, gastric fluid, amniotic fluid, cyst fluid, peritoneal fluid, marrow, bile, other body fluids, tissue obtained from a biopsy, skin, or hair.
[0083] The term “sequence,” as used herein, generally refers to a biological sequence including one-dimensional monomers that can be assembled to generate a polymer. Non-limiting examples of sequences include nucleotide sequences (e.g., ssDNA, dsDNA, andRNA), amino acid sequences (e.g., proteins, peptides, and polypeptides), and carbohydrates.
[0084] The term “subject,” as used herein, refers to a human. A subject can include a healthy or asymptomatic individual, an individual that has or is suspected of having a disease (e.g., NSCLC) or a pre-disposition to the disease, and / or an individual that needs therapy or suspected of needing therapy. A subject can be a patient.
[0085] As used herein, a “model” may include one or more algorithms, one or more mathematical techniques, one or more machine-learning algorithms, or a combination thereof.
[0086] As used herein, “machine-learning” may be the practice of using algorithms to parse data, learn from it, and then make a determination or prediction about something in the world. Machine-learning uses algorithms that can learn from data without relying on rules- based programming. A machine-learning algorithm may include a parametric model, a nonparametric model, a deep learning model, a neural network, a linear discriminant analysis model, a quadratic discriminant analysis model, a support vector machine, a random forest algorithm, a nearest neighbor algorithm, a combined discriminant analysis model, a k-means clustering algorithm, a supervised model, an unsupervised model, logistic regression model, a univariate regularized regression model (e.g., a univariate least absolute shrinkage and selection operator (LASSO) regression model), a multivariate regularized regression model (e.g., a multivariate LASSO regression model), a penalized univariate regularized regression model (e.g., a penalized univariate LASSO regression model), a penalized multivariate regularized regression model (e.g., a penalized multivariate LASSO regression model), or another type of model.
[0087] As used herein, an “artificial neural network (ANN)” or “neural network (NN)” may refer to mathematical algorithms or computational models that mimic an interconnected group of artificial nodes or neurons that processes information based on a connectionistic approach to computation. Neural networks, which may also be referred to as neural nets, can employ one or more layers of nonlinear units to predict an output for a received input. Some neural networks include one or more hidden layers in addition to an output layer. The output of each hidden layer is used as input to the next layer in the network, i.e., the next hiddenlayer or the output layer. Each layer of the network generates an output from a received input in accordance with current values of a respective set of parameters. In the various embodiments, a reference to a “neural network” may be a reference to one or more neural networks.
[0088] A neural network may process information in two ways: when it is being trained it is in training mode and when it puts what it has learned into practice it is in inference (or prediction) mode. Neural networks learn through a feedback process (e.g., backpropagation) which allows the network to adjust the weight factors (modifying its behavior) of the individual nodes in the intermediate hidden layers so that the output matches the outputs of the training data. In other words, a neural network learns by being fed training data (learning examples) and eventually learns how to reach the correct output, even when it is presented with a new range or set of inputs. A neural network may include, for example, without limitation, at least one of a Feedforward Neural Network (FNN), a Recurrent Neural Network (RNN), a Modular Neural Network (MNN), a Convolutional Neural Network (CNN), a Residual Neural Network (ResNet), an Ordinary Differential Equations Neural Networks (neural- ODE), or another type of neural network.
[0089] As used herein, a “target glycopeptide analyte,” may refer to a peptide structure (e.g., glycosylated or aglycosylated / non-glycosylated), a fraction of a peptide structure, a sub-structure (e.g., a glycan or a glycosylation site) of a peptide structure, a product of one or more of the above listed structures and sub-structures, associated detection molecules (e.g., signal molecule, label, or tag), or an amino acid sequence that can be measured by mass spectrometry. For example, a quadrupole mass analyzer of mass spectrometer can be configured to filter a preselected m / z value that corresponds to a target glycopeptide analyte in an ionized state.
[0090] As used herein, a “peptide data set,” may be used interchangeably with “peptide structure data” and can refer to any data of or relating to a peptide presence or abundance. For example, peptide data set or peptide structure data can be based upon a mass spectrometry run, an EEISA, or western blot. A peptide data set can comprise data obtained from a sample or biological sample using mass spectrometry. A peptide dataset can comprise data relating to a non-glycosylated endogenous peptide (NGEP) external standard, data relating to an internal standard, and data relating to a target glycopeptide analyte of a sample.A peptide data set can result from analysis originating from a single run. In some embodiments, the peptide data set can include raw abundance and mass to charge ratios for one or more peptides.
[0091] As used herein, a “non-glycosylated endogenous peptide” (“NGEP”), which may also be referred to as an aglycosylated peptide, may refer to a peptide structure that does not comprise a glycan molecule. In various embodiments, an NGEP and a target glycopeptide analyte can originate from the same subject. In various embodiments, an NGEP can be labeled with an isotope in preparation for mass spectrometry analysis.
[0092] As used herein, a “transition,” may refer to or identify a peptide structure. In some embodiments, a transition can refer to the specific pair of m / z values associated with a precursor ion and a product or fragment ion.
[0093] As used herein, an “abundance value” may refer to “abundance” or a quantitative value associated with abundance.
[0094] As used herein, “abundance,” may refer to a quantitative value generated using mass spectrometry. In various embodiments, the quantitative value may relate to an amount of a particular peptide structure (e.g., biomarker) present in a biological sample. In some embodiments, the amount may be in relation to other structures present in the sample (e.g., relative abundance). In some embodiments, the quantitative value may comprise an amount of an ion produced using mass spectrometry. In some embodiments, the quantitative value may be associated with an m / z value (e.g., abundance on y-axis and m / z on x-axis). In other embodiments, the quantitative value may be expressed in atomic mass units.
[0095] As used herein, “relative abundance,” may refer to a comparison of two or more abundances. In various embodiments, the comparison may comprise comparing one peptide structure to a total number of peptide structures. In some embodiments, the comparison may comprise comparing one peptide glycoform (e.g., two identical peptides differing by one or more glycans) to a set of peptide glycoforms. In some embodiments, the comparison may comprise comparing a number of ions having a particular m / z ratio by a total number of ions detected. In various embodiments, a relative abundance can be expressed as a ratio. In other embodiments, a relative abundance can be expressed as a percentage. Relative abundance can be presented on a y-axis of a mass spectrum plot.
[0096] As used herein, an “internal standard,” may refer to something that can be contained (e.g., spiked-in) in the same sample as a target glycopeptide analyte undergoing mass spectrometry analysis. Internal standards can be used for calibration purposes. Additionally, internal standards can be used in the systems and method described herein. In some aspects, an internal standard can be selected based on similarity m / z and or retention times and can be a “surrogate” if a specific standard is too costly or unavailable. Internal standards can be heavy labeled or non-heavy labeled. In some instances, the term internal standard can be referred to with the abbreviation ISTD.
[0097] “Healthy” or “normal” as used herein refers to an individual who does not have NSCLC. The individual may have other diseases, disorders, and / or conditions, which may or may not relate to lung cancer.
[0098] “Treatment” refers to a therapeutic intervention that ameliorates a sign or symptom of a disease or pathological condition after it has begun to develop. The term “ameliorating,” with reference to a disease or pathological condition, refers to any observable beneficial effect of the treatment. The beneficial effect can be evidenced, for example, by a delayed onset of clinical symptoms of the disease in a susceptible subject, a reduction in severity of some or all clinical symptoms of the disease, a slower progression of the disease, an improvement in the overall health or well-being of the subject, or by other parameters well known in the art that are specific to the particular disease. A “prophylactic” treatment is a treatment administered to a subject who does not exhibit signs of a disease or exhibits only early signs for the purpose of decreasing the risk of developing pathology.II. Exemplary Workflow
[0099] FIG. 1 is a schematic diagram of an exemplary workflow 100 for the detection of peptide structures associated with a disease state for use in diagnosis and / or treatment in accordance with one or more embodiments. Workflow 100 may include various operations including, for example, sample collection 102, sample intake 104, sample preparation and processing 106, data analysis 108, and output generation 130.
[0100] Sample collection 102 may include, for example, obtaining a biological sample 112 of one or more subjects, such as subject 114. Biological sample 112 may take the form of a specimen obtained via one or more sampling methods. Biological sample 112 may berepresentative of subject 114 as a whole or of a specific tissue, cell type, or other category or sub-category of interest. Biological sample 112 may be obtained in any of a number of different ways. In various embodiments, biological sample 112 includes whole blood sample 116 obtained via a blood draw. In other embodiments, biological sample 112 includes set of aliquoted samples 118 that includes, for example, a serum sample, a plasma sample, a blood cell (e.g., white blood cell (WBC), red blood cell (RBC) sample, another type of sample, or a combination thereof. Biological sample 112 may include nucleotides (e.g., ssDNA, dsDNA, RNA), organelles, amino acids, peptides, proteins, carbohydrates, glycoproteins, or any combination thereof.
[0101] In various embodiments, a single run can analyze a sample (e.g., the sample including a peptide analyte), an external standard (e.g., an NGEP of a serum sample), and an internal standard. As such, abundance values (e.g., abundance or raw abundance) for the external standard, the internal standard, and target glycopeptide analyte can be determined by mass spectrometry in the same run.
[0102] In various embodiments, external standards may be analyzed prior to analyzing samples. In various embodiments, the external standards can be run independently between the samples. In some embodiments, external standards can be analyzed after every 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or more experiments. In various embodiments, external standard data can be used in some or all of the normalization systems and methods described herein. In additional embodiments, blank samples may be processed to prevent column fouling.
[0103] Sample intake 104 may include one or more various operations such as, for example, aliquoting, registering, processing, storing, thawing, and / or other types of operations. In one or more embodiments, when biological sample 112 includes whole blood sample 116, sample intake 104 includes aliquoting whole blood sample 116 to form a set of aliquoted samples that can then be sub-aliquoted to form set of samples 120.
[0104] Sample preparation and processing 106 may include, for example, one or more operations to form set of peptide structures 122. In various embodiments, set of peptide structures 122 may include various fragments of unfolded proteins that have undergone digestion and may be ready for analysis.
[0105] Further, sample preparation and processing 106 may include, for example, data acquisition 124 based on set of peptide structures 122. For example, data acquisition 124 may include use of, for example, but is not limited to, a liquid chromatography-mass spectrometry (LC-MS) system.
[0106] Data analysis 108 may include, for example, peptide structure analysis 126. In some embodiments, data analysis 108 also includes output generation 110. Peptide structure analysis can include determining the composition and the associated quantity for the various peptides and glycopeptides present in the sample by processing the output of a mass spectrometer. In other embodiments, output generation 110 may be considered a separate operation from data analysis 108. Output generation 110 may include, for example, generating final output 128 based on the results of peptide structure analysis 126. In various embodiments, final output 128 may be used for determining the research, diagnosis, and / or treatment of a state associated with NSCLC.
[0107] In various embodiments, final output 128 is comprised of one or more outputs. Final output 128 may take various forms. For example, final output 128 may be a report that includes, for example, a diagnosis output, a treatment output (e.g., a treatment design output, a treatment plan output, or combination thereof), analyzed data (e.g., relativized and normalized) or combination thereof. In some embodiments, the report can comprise a target glycopeptide analyte concentration as a function of the NGEP concentration value and the normalized abundance value. In some embodiments, final output 128 may be an alert (e.g., a visual alert, an audible alert, etc.), a notification (e.g., a visual notification, an audible notification, an email notification, etc.), an email output, or a combination thereof. In some embodiments, final output 128 may be sent to remote system 130 for processing. Remote system 130 may include, for example, a computer system, a server, a processor, a cloud computing platform, cloud storage, a laptop, a tablet, a smartphone, some other type of mobile computing device, or a combination thereof.
[0108] In other embodiments, workflow 100 may optionally exclude one or more of the operations described herein and / or may optionally include one or more other steps or operations other than those described herein (e.g., in addition to and / or instead of those described herein). Accordingly, workflow 100 may be implemented in any of a number of different ways for use in the research, diagnosis, and / or treatment of, for example, NSCLC.III. Detection and Quantification of Peptide Structures
[0109] FIG. 2A and FIG. 2B are schematic diagrams of a workflow for sample preparation and processing 106 in accordance with one or more embodiments. FIG. 2 A and FIG. 2B are described with continuing reference to FIG. 1. Sample preparation and processing 106 may include, for example, preparation workflow 200 shown in FIG. 2A and data acquisition 124 shown in FIG. 2B.A. Sample Preparation and Processing
[0110] FIG. 2A is a schematic diagram of preparation workflow 200 in accordance with one or more embodiments. Preparation workflow 200 may be used to prepare a sample, such as a sample of set of samples 120 in FIG. 1, for analysis via data acquisition 124. For example, this analysis may be performed via mass spectrometry (e.g., LC-MS). In various embodiments, preparation workflow 200 may include denaturation and reduction 202, alkylation 204, and digestion 206.
[0111] In general, polymers, such as proteins, in their native form, can fold to include secondary, tertiary, and / or other higher order structures. Such higher order structures may functionalize proteins to complete tasks (e.g., enable enzymatic activity) in a subject.Further, such higher order structures of polymers may be maintained via various interactions between side chains of amino acids within the polymers. Such interactions can include ionic bonding, hydrophobic interactions, hydrogen bonding, and disulfide linkages between cysteine residues. However, when using analytic systems and methods, including mass spectrometry, unfolding such polymers (e.g., peptide / protein molecules) may be desired to obtain sequence information. In some embodiments, unfolding a polymer may include denaturing the polymer, which may include, for example, linearizing the polymer.
[0112] In one or more embodiments, denaturation and reduction 202 can be used to disrupt higher order structures (e.g., secondary, tertiary, quaternary, etc.) of one or more proteins (e.g., polypeptides and peptides) in a sample (e.g., one of set of samples 120 in FIG. 1). Denaturation and reduction 202 includes, for example, a denaturation procedure and a reduction procedure. In some embodiments, the denaturation procedure may be performed using, for example, thermal denaturation, where heat is used as a denaturing agent (e.g.,heating the sample to about 90°C to about 100 °C for about 1 to about 10 minutes. The thermal denaturation can disrupt ionic bonding, hydrophobic interactions, and / or hydrogen bonding.
[0113] In one or more embodiments, the denaturation procedure may include using one or more denaturing agents, temperature (e.g., heat), or both. These one or more denaturing agents may include, for example, but are not limited to, any number of chaotropic salts (e.g., urea, guanidine), surfactants (e.g., sodium dodecyl sulfate (SDS), beta octyl glucoside, Triton X-100), or combination thereof. In some cases, such denaturing agents may be used in combination with heat when sample preparation workflow further includes a cleanup procedure.
[0114] The resulting one or more denatured (e.g., unfolded, linearized) proteins may then undergo further processing in preparation of analysis. For example, a reduction procedure may be performed in which one or more reducing agents are applied. In various embodiments, a reducing agent can produce an alkaline pH. A reducing agent may take the form of, for example, without limitation, dithiothreitol (DTT), tris(2-carboxyethyl)phosphine (TCEP), or some other reducing agent. The reducing agent may reduce (e.g., cleave) the disulfide linkages between cysteine residues of the one or more denatured proteins to form one or more reduced proteins.
[0115] In various embodiments, the one or more reduced proteins resulting from denaturation and reduction 202 may undergo a process to prevent the reformation of disulfide linkages between, for example, the cysteine residues of the one or more reduced proteins. This process may be implemented using alkylation 204 to form one or more alkylated proteins. For example, alkylation 204 may be used to add an acetamide group to a sulfur on each cysteine residue to prevent disulfide linkages from reforming. In various embodiments, an acetamide group can be added by reacting one or more alkylating agents with a reduced protein. The acetamide group or alkylation group that attaches to the protein or peptide results in a different form that is not naturally occurring in nature. The one or more alkylating agents may include, for example, one or more acetamide salts. An alkylating agent may take the form of, for example, iodoacetamide (IAA), 2-chloroacetamide, some other type of acetamide salt, or some other type of alkylating agent.
[0116] In some embodiments, alkylation 204 may include a quenching procedure. The quenching procedure may be performed using one or more reducing agents (e.g., one or more of the reducing agents described above).
[0117] In various embodiments, the one or more alkylated proteins formed via alkylation 204 can then undergo digestion 206 in preparation for analysis (e.g., mass spectrometry analysis). Digestion 206 of a protein may include cleaving the protein at or around one or more cleavage sites (e.g., site 205 which may be one or more amino acid residues). For example, without limitation, an alkylated protein may be cleaved at the carboxyl side of lysine or arginine residues. This type of cleavage may break the protein into various segments, which include one or more peptide structures (e.g., glycosylated or aglycosylated).
[0118] In various embodiments, digestion 206 is performed using one or more proteolysis catalysts. For example, an enzyme can be used in digestion 206. In some embodiments, the enzyme takes the form of trypsin. In other embodiments, one or more other types of enzymes (e.g., proteases) may be used in addition to or in place of trypsin. These one or more other enzymes include, but are not limited to, LysC, LysN, AspN, GluC, and ArgC. In some embodiments, digestion 206 may be performed using tosyl phenylalanyl chloromethyl ketone (TPCK)-treated trypsin, one or more engineered forms of trypsin, one or more other formulations of trypsin, or a combination thereof. In some embodiments, digestion 206 may be performed in multiple steps, with each involving the use of one or more digestion agents. For example, a secondary digestion, tertiary digestion, etc. may be performed. In one or more embodiments, trypsin is used to digest serum samples. In one or more embodiments, trypsin / LysC cocktails are used to digest plasma samples.
[0119] In some embodiments, digestion 206 further includes a quenching procedure. The quenching procedure may be performed by acidifying the sample (e.g., to a pH <3). In some embodiments, formic acid may be used to perform this acidification.
[0120] In various embodiments, preparation workflow 200 further includes post-digestion procedure 207. Post-digestion procedure 207 may include, for example, a cleanup procedure. The cleanup procedure may include, for example, the removal of unwanted components in the sample that results from digestion 206. For example, unwanted components may include, but are not limited to, inorganic ions, surfactants, etc. In some embodiments, post-digestionprocedure 207 further includes a procedure for the addition of heavy-labeled peptide internal standards.
[0121] Although preparation workflow 200 has been described with respect to a sample created or taken from biological sample 112, such as a blood-based sample 116 (e.g., a whole blood sample, a plasma sample, a serum sample, etc.), sample preparation workflow 200 may be similarly implemented for other types of samples (e.g., tears, urine, tissue, interstitial fluids, sputum, etc.) to produce set of peptide structures 122.B. Peptide Structure Identification and Quantification
[0122] FIG. 2B is a schematic diagram of data acquisition 124 in accordance with one or more embodiments. In various embodiments, data acquisition 124 can commence following sample preparation 200 described in FIG. 2A. In various embodiments, data acquisition 124 can comprise quantification 208, quality control 210, and peak integration and normalization 212.
[0123] In various embodiments, quantification 208 of peptides and glycopeptides can incorporate use of liquid chromatography -mass spectrometry LC-MS instrumentation. For example, LC-MS / MS, or tandem MS may be used. In general, LC-MS (e.g., LC-MS / MS) can combine the physical separation capabilities of liquid chromatography (LC) with the mass analysis capabilities of mass spectrometry (MS). According to some embodiments described herein, this technique allows for the separation of digested peptides to be fed from the LC column into the MS ion source through an interface. In various embodiments, quantification 208 is targeted quantification.
[0124] In various embodiments, any LC-MS device can be incorporated into the workflow described herein. In various embodiments, an instrument or instrument system suited for identification and quantification 208 may include, for example, a Triple Quadrupole LC-MS. In various embodiments, quantification 208 is performed using multiple reaction monitoring mass spectrometry (MRM-MS). MRM is a mass spectrometry method in which a precursor ion of a particular m / z (e.g., peptide analyte) is selected in the first quadrupole (QI) and transmitted to the second quadrupole (Q2) for fragmentation. The resulting product ions are then transmitted to the third quadrupole (Q3), which detects only product ions with selected predefined m / z values. The particular m / z value set for the firstquadrupole (QI) and the selected predefined m / z values of the third quadrupole have a mass range that ranges within + / - 1, + / - 0.5, or + / -0.1 m / z values.
[0125] In various embodiments described herein, identification of a particular protein or peptide and an associated quantity can be assessed. In various embodiments described herein, identification of a particular glycopeptide and an associated quantity can be assessed. In various embodiments described herein, identification of a particular glycan and an associated quantity can be assessed. In various embodiments described herein, particular glycans can be matched to a glycosylation site on a protein or peptide and the abundance values measured. In various embodiments, a glycopeptide of any of SEQ ID NOs: 27-99 and an associated quantity is assessed. In various embodiments, a glycopeptide provided in Table 3 and an associated quantity is assessed. The glycan portion of the glycopeptide is provided in Tables 4A and 4B that indicate the corresponding symbol structure and composition for the glycopeptides of Table 3.
[0126] In some cases, quantification 208 includes using a specific collision energy associated for the appropriate fragmentation to consistently see an abundant product ion. Glycopeptides may have a lower collision energy than aglycosylated peptide structures. When analyzing a sample that includes glycopeptides, the source voltage and gas temperature may be lowered as compared to generic proteomic analysis.
[0127] In various embodiments, quality control 210 procedures can be put in place to optimize data quality. In various embodiments, measures can be put in place allowing only errors within acceptable ranges outside of an expected value. In various embodiments, employing statistical models (e.g., using Westgard rules) can assist in quality control 210. For example, quality control 210 may include, for example, assessing the retention time and abundance of representative peptide structures (e.g., glycosylated and / or aglycosylated) and spiked-in internal standards, in either every sample, or in each quality control sample (e.g., pooled serum digest).
[0128] Peak integration and normalization 212 may be performed to process the data that has been generated and transform the data into a format for analysis. For example, peak integration and normalization 212 may include converting abundance data for various product ions that were detected for a selected peptide structure into a single quantificationmetric (e.g., a relative quantity, an adjusted quantity, a normalized quantity, a relative concentration, an adjusted concentration, a normalized concentration, etc.) for that peptide structure. In some embodiments, peak integration and normalization 212 may be performed using one or more of the techniques described in U.S. Patent Publication No.2020 / 0372973 Al and / or US Patent Publication No. 2020 / 0240996A1, the disclosures of which are incorporated by reference herein in their entireties
[0129] In some embodiments, the presence, absence, and / or amount of at least one peptide structures is determined by a method other than mass spectrometry, for example by ELISA or immunoblotting (such as western blot). In some embodiments, the presence, absence / and or amount of a peptide structure set forth in Table 3 and / or Table 8 is determined by a method other than mass spectrometry, for example by ELISA or immunoblotting (such as western blot). In some embodiments, the presence, absence, and / or amount of a peptide structure comprising a sequence set forth in SEQ ID NOs: 27-99 is determined by a method other than mass spectrometry, for example by ELISA or immunoblotting (such as western blot).IV. Computer Implemented Systems
[0130] FIG. 3 is a block diagram of an analysis system 300, in accordance with the presently disclosed embodiments. For example, in accordance with the presently disclosed embodiments, the analysis system 300 may include any computing platform that may be utilized for performing one or more methods of classifying a biological sample obtained from a subject with respect to a plurality of states associated with NSCLC, detecting the presence of one of a plurality of states associated with NSCLC, treating NSCLC in a subject, determining techniques for treating NSCLC in a subject, diagnosing an individual with NSCLC, and training a model to diagnose a subject with one of a plurality of states associated with NSCLC, in accordance with the presently disclosed embodiments. Analysis system 300 can be used to detect and analyze various peptide structures that have been associated with various states of NSCLC. Analysis system 300 is one example of an implementation for a system that may be used to perform data analysis 108. Analysis system 300 may include computing platform 302 and data store 304.
[0131] In certain embodiments, analysis system 300 may also include display system 306. Computing platform 302 may take various forms. In certain embodiments, computing platform 302 may include a single computer (or computer system) or multiple computers in communication with each other. In other examples, computing platform 302 takes the form of a cloud computing platform. Data store 304 and display system 306 may each be in communication with computing platform 302. In some examples, data store 304, display system 306, or both may be considered part of or otherwise integrated with computing platform 302. Thus, in some examples, computing platform 302, data store 304, and display system 306 may be separate components in communication with each other, but in other examples, some combination of these components may be integrated together.Communication between these different components may be implemented using any number of wired communications links, wireless communications links, optical communications links, or a combination thereof.
[0132] In certain embodiments, analysis system 300 may include, for example, peptide structure analyzer 308, which may be implemented using hardware, software, firmware, or a combination thereof. In certain embodiments, peptide structure analyzer 308 is implemented using computing platform 302. Peptide structure analyzer 308 receives peptide structure data 310 for processing. Peptide structure data 310 may be, for example, the peptide structure data that is output from sample preparation and processing 106 in FIG. 1, FIG. 2A, and FIG. 2B. Accordingly, peptide structure data 310 may correspond to set of peptide structures 122 identified for biological sample 112 and may thereby correspond to biological sample 112. Peptide structure data 310 can be sent as input into peptide structure analyzer 308, retrieved from data store 304 or some other type of storage (e.g., cloud storage), accessed from cloud storage, or obtained in some other manner. In some cases, peptide structure data 310 may be retrieved from data store 304 in response to (e.g., directly or indirectly based on) receiving user input entered by a user via an input device. Peptide structure data 310 may include quantification data for the plurality of peptide structures. For example, peptide structure data 310 may include a set of quantification metrics for each peptide structure of a plurality of peptide structures. A quantification metric for a peptide structure may be selected as one of a relative quantity, an adjusted quantity, a normalized quantity, a relative abundance, an adjusted abundance, and a normalized abundance. In some cases, a quantification metric for a peptide structure is selected from one of a relative concentration, an adjusted concentration,and a normalized concentration. In this manner, peptide structure data 310 may provide abundance information about the plurality of peptide structures with respect to biological sample 112.
[0133] In certain embodiments, a peptide structure of set of peptide structures 312 may include a glycosylated peptide structure, or glycopeptide structure, that is defined by a peptide sequence and a glycan structure attached to a linking site of the peptide sequence. For example, the peptide structure may be a glycopeptide or a portion of a glycopeptide. In certain embodiments, a peptide structure of set of peptide structures 312 may include an aglycosylated peptide structure that is defined by a peptide sequence. For example, the peptide structure may be a tag glycopeptide or a portion of a tag glycopeptide and may be referred to as a quantification peptide. A tag peptide can be a peptide with at least one isotopically labeled amino acid. Set of peptide structures 312 may be identified as being those most predictive or relevant to the symptomatic disease state based on training of model 314. In certain embodiments, set of peptide structures 312 may include at least one, at least two, at least three, at least four, at least five, at least 10, at least 15, at least 20, at least 25, or all of the peptide structures identified in Table 3 below. The number of peptide structures selected from Table 3 for inclusion in set of peptide structures 312 may be based on, for example, a desired level of accuracy. In certain embodiments, an N number of peptide structures may be selected from Table 3 for inclusion in set of peptide structures 312, in which N is an integer from 1-73. In certain embodiments, set of peptide structures 312 may include at least one, at least two, at least three, at least four, at least five, at least 10, at least 15, or all of the peptide structures identified in Table 8 below. The number of peptide structures selected from Table 8 for inclusion in set of peptide structures 312 may be based on, for example, a desired level of accuracy. In certain embodiments, an N number of peptide structures may be selected from Table 3 for inclusion in set of peptide structures 312, in which N is an integer from 1-19.
[0134] Peptide structure analyzer 308 may include model 314 that may be able to receive peptide structure data 310 for processing. Model 314 may be implemented in any of a number of different ways. Model 314 may be implemented using any number of models, functions, equations, algorithms, and / or other mathematical techniques. In certain embodiments, model 314 may include one or more machine-learning systems 316, which may include any number of machine-learning models and / or algorithms. For example, machine-learning systems 316 may include, without limitation, at least one of a parametricmodel, a non-parametric model, deep learning model, a neural network, a linear discriminant analysis model, a quadratic discriminant analysis model, a support vector machine, a random forest algorithm, a nearest neighbor algorithm (e.g., a k-Nearest Neighbors algorithm), a combined discriminant analysis model, a k- means clustering algorithm, a supervised machine-learning model, an unsupervised machine-learning model, a logistic regression model, a univariate regularized regression model (e.g., a univariate least absolute shrinkage and selection operator (LASSO) regression model), a multivariate regularized regression model (e.g., a multivariate LASSO regression model), a penalized univariate regularized regression model (e.g., a penalized univariate LASSO regression model), a penalized multivariate regularized regression model (e.g., a penalized multivariate LASSO regression model), or any of various other models that may be utilized to predict the presence and / or absence of NSCLC in a subject, in accordance with the presently-disclosed embodiments. In certain embodiments, model 314 may include machine-learning systems 316, which may include any number of or combination of the models or algorithms described above.
[0135] In certain embodiments, model 314 analyzes the portion (e.g., some or all of) peptide structure data 310 corresponding set of peptide structures 312 to generate disease indicator 318 that classifies biological sample 112 as evidencing a corresponding state of a plurality of states 320 associated with NSCLC. Disease indicator 318 may take various forms. In certain embodiments, disease indicator 318 is a score that indicates a classification of the corresponding state for biological sample 112. In certain embodiments, the classification of the corresponding state for biological sample 112 may be a binary classification (e.g., a prediction score of 0.8, 0.9, or otherwise approximately “1.0” may indicate an NSCLC state, while a prediction score of O.L, 0.2. , otherwise approximately “0.0” may indicate a healthy state). Thus, in accordance with the presently-disclosed embodiments, the disease indicator 318 may simply indicate whether a subject has NSCLC or does not have NSCLC based on the peptide structure data 310.
[0136] In certain embodiments, disease indicator 318 may include a score that indicates a probability that a subject (e.g., subject 114 in FIG. 1) falls within one of the states 320 associated with NSCLC. For example, disease indicator 318 may include one or more scores, each of which may indicate whether biological sample 112 has one of the corresponding states 320 associated with NSCLC. In some examples, disease indicator 318 may include a score for each of the states 320 associated with NSCLC. For example, as generally notedabove, the states 320 associated with NSCLC may include a binary classification, in which, for example, the machine-learning systems 316 may generate prediction value of 0.8, 0.9, or otherwise approximately “1.0” to indicate the presence of NSCLC and may generate prediction value of 0.1, 0.2, otherwise approximately “0.0” to indicate the absence of NSCLC (e.g., a healthy). In certain embodiments, machine-learning systems 316 may include regularized regression model (e.g., a univariate regularized regression model or a multivariate regularized regression model). For example, in one embodiment, the regularized regression may include a LASSO regression model (e.g., a univariate LASSO regression model or a multivariate LASSO regression model) trained to compute disease indicator 318. In another embodiment, the regularized regression model may be trained to, for example, classify a biological sample obtained from a subject with respect to a plurality of states associated with NSCLC, detect the presence of one of a plurality of states associated with NSCLC; determine techniques for treating NSCLC in a subject; and determine techniques for diagnosing an individual with NSCLC, in accordance with the presently disclosed embodiments. The regularized regression model may be trained to identify weight coefficients for peptide structures of set of peptide structures 312. Peptide structure analyzer 308 may generate final output 128 based on disease indicator 318 that is output by model 314. In other embodiments, final output 128 may be an output generated by model 314.
[0137] In certain embodiments, final output 128 may include disease indicator 318. In other embodiments, final output 128 may include diagnosis output 324 and / or treatment output 326. Diagnosis output 324 may include, for example, an identification of a classification of which of the states 320 evidenced by biological sample 112 based on disease indicator 318. Treatment output 326 may include, for example, at least one of an identification of a therapy to treat the subject, a design for the therapy, or a treatment plan for administering the therapy. In certain embodiments, the therapy is a surgery, a chemotherapeutic therapy, a patient- specific therapy, a targeted immunotherapy, a radiation procedure, a radiofrequency ablation (RFA) procedure, or a combination thereof. Final output 128 may be sent to remote system 130 for processing in some examples. In other embodiments, final output 128 may be displayed on graphical user interface 328 in display system 306 for viewing by a human operator. The human operator may use final output 128 to diagnose and / or treat the subject when final output 128 indicates the subject has NSCLC.
[0138] FIG. 4 illustrates a block diagram of a computer system that may be utilized for performing one or more methods of classifying a biological sample obtained from a subject with respect to a plurality of states associated with NSCLC, detecting the presence of one of a plurality of states associated with NSCLC, treating NSCLC in a subject, treating NSCLC in a subject, diagnosing an individual with NSCLC, and training a model to diagnose a subject with one of a plurality of states associated with NSCLC, in accordance with the presently disclosed embodiments. Computer system 400 may be an example of one implementation for computing platform 302 described above in FIG. 3. In certain embodiments, computer system 400 can include a bus 402 or other communication mechanism for communicating information, and a processor 404 coupled with bus 402 for processing information. In certain embodiments, computer system 400 can also include a memory, which can be a randomaccess memory (RAM) 406 or other dynamic storage device, coupled to bus 402 for determining instructions to be executed by processor 404. Memory also can be used for storing temporary variables or other intermediate information during execution of instructions to be executed by processor 404. In various embodiments, computer system 400 can further include a read only memory (ROM) 408 or other static storage device coupled to bus 402 for storing static information and instructions for processor 404. A storage device 410, such as a magnetic disk or optical disk, can be provided and coupled to bus 402 for storing information and instructions.
[0139] In certain embodiments, computer system 400 can be coupled via bus 402 to a display 412, such as a cathode ray tube (CRT) or liquid crystal display (LCD), for displaying information to a computer user. An input device 414, including alphanumeric and other keys, can be coupled to bus 402 for communicating information and command selections to processor 404. Another type of user input device is a cursor control 416, such as a mouse, a joystick, a trackball, a gesture input device, a gaze -based input device, or cursor direction keys for communicating direction information and command selections to processor 404 and for controlling cursor movement on display 412. This input device 414 typically has two degrees of freedom in two axes, a first axis (e.g., x) and a second axis (e.g., y), that allows the device to specify positions in a plane. However, it should be understood that input devices 414 allowing for three-dimensional (e.g., x, y, and z) cursor movement are also contemplated herein.V. Methods of DiagnosingA. General Methodology
[0140] In some embodiments, the methods provided herein are useful for diagnosing NSCLC. In some embodiments, the method is useful for diagnosing early-stage NSCLC. In some embodiments, the method is useful for diagnosing late-stage NSCLC. In some embodiments, the method comprises diagnosing an individual with NSCLC, wherein the individual has stage 0, stage I, stage II, stage III, or stage IV NSCLC. In some embodiments, the method comprises diagnosing an individual with NSCLC before the onset of disease symptoms and pathological conditions of NSCLC. In some embodiments, the method comprises diagnosing an individual with NSCLC after the onset of disease symptoms and pathological conditions of NSCLC. In some embodiments, the method comprises diagnosing an individual with NSCLC during disease symptoms and pathological conditions of NSCLC. In some embodiments, the method comprises diagnosing an individual with NSCLC before disease progression. In some embodiments, the method comprising diagnosing an individual with NSCLC before metastasis.
[0141] In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structures provided in Table 3 and / or Table 8. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the glycopeptide structure data comprises one or more glycopeptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC, based upon one or more peptide structure provided in Table 3 or Table 8, and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC, based upon one or more peptidestructure provided in Table 3 or Table 8, and administering a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC, based upon one or more glycopeptide structure provided in Table 3 or Table 8, and administering a treatment for NSCLC based upon the classification.
[0142] In some embodiments, the diagnosis is based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, at least ten, at least 15, at least 20, or at least 25 peptide structures from Table 3. In some embodiments, the diagnosis is based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, at least ten, at least 15, at least 20, or at least 25 glycopeptides from Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptide comprising the amino acid sequence of SEQ ID NOs: 27-60 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of two or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of three or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of four or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of five or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 10 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In someembodiments, the diagnosis is based upon the presence and / or amount of 15 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 20 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 25 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structure provided in Table 3 and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structure provided in Table 3 and administering a treatment for NSCLC based upon the classification.
[0143] In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of two or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of three or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of four or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. Insome embodiments, the diagnosis is based upon the presence and / or amount of five or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 10 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 15 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 20 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 25 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 3 and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 3 and administering a treatment for NSCLC based upon the classification.
[0144] In some embodiments, the method comprises determining whether an individual has NSCLC by inputting quantification data from peptide and / or glycopeptide structure data for a set of peptide structures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, 20 or more, 25 or more, or each of the peptides and / or glycopeptide consisting of the amino acid sequence of SEQ ID NOs: 27-60 into a machine-learning model trained to identify a disease indicator, identifying by the machinelearning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the disease indicator comprises one or more scores that indicate a probabilitythat the subject falls within one of the states associated with NSCLC (e.g., having NSCLC or not having NSCLC). In some embodiments, the machine-learning model is a regularized regression model. In some embodiments, the regularized model comprises a LASSO regression model. In some embodiments, the peptide structures are detected using LC-MS. In some embodiments, the LC-MS comprises LC-MS / MS and LC-MS / MS running in a MRM mode. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the classification described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the classification described herein.
[0145] In some embodiments, the diagnosis is based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, at least 10, at least 15, at least 20, or at least 25 peptide structures from Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptide comprising the amino acid sequence of SEQ ID NOs: 61-99 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptide comprising the amino acid sequence of SEQ ID NOs: 61-99 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of two or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of three or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of four or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of five or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. Insome embodiments, the diagnosis is based upon the presence and / or amount of 10 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 15 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 20 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 25 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 3 and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 3 and administering a treatment for NSCLC based upon the classification.
[0146] In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of two or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of three or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associatedglycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of four or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of five or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 10 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 15 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 20 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 25 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 3 and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 3 and administering a treatment for NSCLC based upon the classification.
[0147] In some embodiments, the method comprises determining whether an individual has NSCLC or is likely to develop NSCLC by inputting quantification data from peptide and / or glycopeptide structure data for a set of peptide structures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, 20 or more, 25 or more, or each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 61-99 into a machine-learning model trained to identify a disease indicator,identifying by the machine-learning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the disease indicator comprises one or more scores that indicate a probability that the subject falls within one of the states associated with NSCLC (e.g., having NSCLC or not having NSCLC). In some embodiments, the machinelearning model is a regularized regression model. In some embodiments, the regularized regression model comprises a LASSO regression model. In some embodiments, the peptide structures are detected using LC-MS. In some embodiments, the LC-MS comprises LC- MS / MS and LC-MS / MS running in a MRM mode. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the classification described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the classification described herein.
[0148] In some embodiments, the diagnosis is based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, at least 10, or at least 15 peptide structures from Table 8. In some embodiments, the diagnosis is based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, at least 10, or at least 15 glycopeptides from Table 8. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptide comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 set forth in Table 8. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of two or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of three or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associatedglycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of four or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of five or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 10 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 15 or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 8 and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 8 and administering a treatment for NSCLC based upon the classification.
[0149] In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 set forth in Table 8. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 set forth in Table 8. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of two or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of three or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34- 37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of four or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34- 37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of five or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34- 37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 10 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of 15 or more peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 3 and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide and / or glycopeptide structure provided in Table 3 and administering a treatment for NSCLC based upon the classification.
[0150] In some embodiments, the method comprises determining whether an individual has NSCLC by inputting quantification data from peptide and / or glycopeptide structure data for a set of peptide structures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, or each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 into a machine-learning model trained to identify a disease indicator, identifying by themachine-learning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the disease indicator comprises one or more scores that indicate a probability that the subject falls within one of the states associated with NSCLC (e.g., having NSCLC or not having NSCLC). In some embodiments, the machine-learning model is a regularized regression model. In some embodiments, the regularized regression model comprises a LASSO regression model. In some embodiments, the peptide structures are detected using LC-MS. In some embodiments, the LC-MS comprises LC-MS / MS and LC- MS / MS running in a MRM mode. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the classification described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the classification described herein.
[0151] In some embodiments, the method comprises determining whether an individual has NSCLC by inputting quantification data from peptide and / or glycopeptide structure data for a set of peptide structures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, or each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-99 into a machine-learning model trained to identify a disease indicator, identifying by the machine-learning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the peptide structures are detected using LC-MS. In some embodiments, the LC-MS comprises LC-MS / MS and LC-MS / MS running in a MRM mode. In some embodiments, the quantification data from peptide and / or glycopeptide comprises a fold-change in peptide and / or glycopeptide abundance for different sample conditions (e.g., NSCLC vs healthy control sample). In some embodiments, the fold-change in peptide and / or glycopeptide abundance is greater than 1, wherein the peptide and / or glycopeptide abundance is increased in the NSCLC sample compared to healthy control sample. In some embodiments, the fold-change in peptide and / or glycopeptide abundance is less than 1, wherein the peptide and / or glycopeptide abundance is decreased in the NSCLC sample compared to healthy control sample. In some embodiments, the fold-change in peptide and / or glycopeptide abundance is equal to 1, wherein the peptide and / or glycopeptide abundance is the sample in the NSCLC sample and the healthy control sample. In some embodiments, the fold-change in peptide and / or glycopeptide abundance isabout 1.0, about 1.2, about 1.4, about 1.6, about 1.8, about 2.0, about 2.2, about 2.4, about 2.6, about 2.8, about 3.0, about 3.5, about 4, about 5, about 6, about 7, about 8, about 9, or about 10 for the NSCLC sample compared to the healthy control sample. In some embodiments, the fold-change in peptide and / or glycopeptide abundance is more than 1.0, more than 1.2, more than 1.4, more than 1.6, more than 1.8, more than 2.0, more than 2.2, more than 2.4, more than 2.6, more than 2.8, more than 3.0, more than 3.5, more than 4, more than 5, more than 6, more than 7, more than 8, more than 9, or more than 10 for the NSCLC sample compared to the healthy control sample. In some embodiments, the fold-change in peptide and / or glycopeptide abundance is about 0.9, about 0.8, about 0.7, about 0.6, about 0.5, about 0.4, about 0.3, about 0.2, or about 0.1 for the NSCLC sample compared to the healthy control sample. In some embodiments, the fold-change in peptide and / or glycopeptide abundance is less than 0.9, less than 0.8, less than 0.7, less than 0.6, less than 0.5, less than 0.4, less than 0.3, less than 0.2, or less than 0.1 for the NSCLC sample compared to the healthy control sample. In some embodiments, the fold-change in peptide and / or glycopeptide abundance is provided for the amino acid sequence of SEQ ID NOs: 27- 99 as set forth by Table 6A and Table 6B. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the quantification data described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the quantification data described herein. In some embodiments, the diagnostic methods herein comprise selecting one or more glycopeptide based upon a fold change cutoff between healthy control samples and NSCLC samples. In some embodiments, the selected glycopeptides are used to train a model to diagnose NSCLC.
[0152] In some embodiments, the method comprises determining whether an individual has NSCLC by inputting quantification data from peptide and / or glycopeptide structure data for a set of peptide structures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, or each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 into a machine-learning model, wherein the machine-learning model comprises a regularized regression model (e.g., LASSO regression model) trained using one or more of a first set of peptide and / or glycopeptide structure coefficients set forth in Table 7 for all stages of NSCLC (stages 1-4), a second set of peptide and / or glycopeptide structure coefficients set forth in Table 7 for early-stage NSCLC (stages 1-2), and a third set of peptide structure coefficients set forth in Table 7 for late- stageNSCLC (stages 3-4). In some embodiments, the first set of peptide structure coefficients used in training the model comprises the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58. In some embodiments, the second set of peptide structure coefficients used in training the model comprises the amino acid sequence of SEQ ID NOs: 27, 30, 36, 41, 43-45, 47, 50-51, 53-58. In some embodiments, the third set of peptide structure coefficients used in training the model comprise the amino acid sequence of SEQ ID NOs: 28-29, 31-33, 36-40, 42, 44-47, 49, 57-60. In some embodiments, the machine-learning model is trained using any one of the set of peptide structure coefficients described herein to identify a disease indicator, identifying by the machine-learning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the disease indicator comprises one or more scores that indicate a probability that the subject falls within one of the states associated with NSCLC (e.g., having NSCLC or not having NSCLC). In some embodiments, the machine-learning model trained using the first set of peptide structure coefficients is a regularized regression model (e.g., LASSO regression model). In some embodiments, the machine-learning model trained using the second set of peptide structure coefficients is a regularized regression model (e.g., LASSO regression model). In some embodiments, the machine-learning model trained using the third set of peptide structure coefficients is a regularized regression model (e.g., LASSO regression model). In some embodiments, the machine-learning model trained using the first set of peptide structure coefficients comprises a regularized regression model (e.g., LASSO regression model). In some embodiments, the machine-learning model trained using the second set of peptide structure coefficients comprises a regularized regression (e.g., LASSO regression model). In some embodiments, the machine-learning model trained using the third set of peptide structure coefficients comprises a regularized regression model (e.g., LASSO regression model). In some embodiments, the peptide structures are detected using LC-MS. In some embodiments, the LC-MS comprises LC-MS / MS and LC-MS / MS running in a MRM mode. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the classification described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the classification described herein.
[0153] In some embodiments, the method comprises determining whether an individual has NSCLC by inputting quantification data from peptide structure data for a set of peptidestructures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, or each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 into a machine-learning model, wherein the machinelearning model consists of a regularized regression model (e.g., LASSO regression model) trained using a first set of peptide structure coefficients set forth in Table 7 for all stages of NSCLC (stages 1-4). In some embodiments, the first set of peptide structure coefficients used in training the model consist of the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58. In some embodiments, the model is trained using the first set of peptide structure coefficients to identify a disease indicator, identifying by the machine-learning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the disease indicator comprises one or more scores that indicate a probability that the subject falls within one of the states associated with NSCLC (e.g., having NSCLC or not having NSCLC). In some embodiments, the machine-learning model trained using the first set of peptide structure coefficients is a regularized regression model. In some embodiments, the regularized regression model trained using the first set of peptide structure coefficients consists of a LASSO regression model. In some embodiments, the peptide structures are detected using LC-MS. In some embodiments, the LC-MS comprises LC-MS / MS and LC- MS / MS running in a MRM mode. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the classification described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the classification described herein.
[0154] In some embodiments, the method comprises determining whether an individual has NSCLC by inputting quantification data from peptide structure data for a set of peptide structures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, or each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 into a machine-learning model, wherein the machinelearning model consists of a regularized regression model (e.g., LASSO regression model) trained using a second set of peptide structure coefficients set forth in Table 7 for early-stage NSCLC (stages 1-2). In some embodiments, the second set of peptide structure coefficients used in training the machine-learning model consist of the amino acid sequence of SEQ ID NOs: 27, 30, 36, 41, 43-45, 47, 50-51, 53-58. In some embodiments, the machine-learningmodel is trained using the second set of peptide structure coefficients to identify a disease indicator, identifying by the machine-learning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the disease indicator comprises one or more scores that indicate a probability that the subject falls within one of the states associated with NSCLC (e.g., having NSCLC or not having NSCLC). In some embodiments, the machinelearning model trained using the second set of peptide structure coefficients is a regularized regression model. In some embodiments, the regularized regression model trained using the second set of peptide structure coefficients consists of a LASSO regression model. In some embodiments, the peptide structure are detected using LC-MS. In some embodiments, the LC-MS comprises LC-MS / MS and LC-MS / MS running in a MRM mode. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the classification described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the classification described herein.
[0155] In some embodiments, the method comprises determining whether an individual has NSCLC by inputting quantification data from peptide structure data for a set of peptide structures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, or each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60 into a machine-learning model, wherein the machinelearning model consists of a regularized regression model (e.g., LASSO regression model) trained using a third set of peptide structure coefficients set forth in Table 7 for late-stage NSCLC (stages 3-4). In some embodiments, the third set of peptide structure coefficients used in training the model consist of the amino acid sequence of SEQ ID NOs: 28-29, 31-33, 36-40, 42, 44-47, 49, 57-60. In some embodiments, the third set of peptide structure coefficients used in training the model consist of the amino acid sequence of SEQ ID NOs: 28-29, 31-33, 36-40, 42, 44-47, 49, 57-60. In some embodiments, the model is trained using the third set of peptide structure coefficients to identify a disease indicator, identifying by the machine-learning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the disease indicator comprises one or more scores that indicate a probability that the subject falls within one of the states associated with NSCLC (e.g., having NSCLC or not having NSCLC). In some embodiments, the machine-learning model trainedusing the third set of peptide structure coefficients is a regularized regression model. In some embodiments, the regularized regression model trained using the third set of peptide structure coefficients consists of a LASSO regression model. In some embodiments, the peptide structure are detected using LC-MS. In some embodiments, the LC-MS comprises LC- MS / MS and LC-MS / MS running in a MRM mode. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the classification described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the classification described herein.
[0156] In some embodiments, the method further comprises collecting a biological sample. In some embodiments, the biological sample comprises a blood sample, a serum sample, a buccal epithelia sample, a nasal epithelia sample, a sputum cytology sample, and a bronchial biopsy sample. In some embodiments, the biological sample comprises a pretreatment peripheral blood sample. In some embodiments, the biological sample is collected from an individual with early-stage NSCLC. In some embodiments, the biological sample is collected from an individual with late-stage NSCLC. In some embodiments, the biological sample is collected from an individual with stage 0, stage I, stage II, stage III, or stage IV NSCLC. In some embodiments, the biological sample is from a healthy individual (i.e., an individual who does not have NSCLC). In some embodiments, the biological sample is collected from an individual before the onset of disease symptoms and pathological conditions of NSCLC. In some embodiments, the biological sample is collected from an individual after the onset of disease symptoms and pathological conditions of NSCLC. In some embodiments, the biological sample is collected from an individual during disease symptoms and pathological conditions of NSCLC. In some embodiments, the biological sample is collected from an individual before disease progression of NSCLC.
[0157] For example, in certain embodiments, the presence or amount of the at least one peptide and / or glycopeptide structure is detected using liquid chromatography-tandem mass spectrometry (LC-MS / MS) running in a multiple reaction monitoring (MRM) mode, or an ELISA. In one embodiment, the at least one peptide and / or glycopeptide structure is none, or below a detection limit. In one embodiment, the at least one peptide and / or glycopeptide structure that is below a detection limit is enriched and / or concentrated in a biological sample using an antibody specific to the at least one peptide and / or glycopeptide structure. In one embodiment, the at least one peptide and / or glycopeptide structure that is below a detectionlimit is enriched and / or concentrated in a biological sample using a protein such as a lectin that binds to a carbohydrate portion of at least one glycopeptide structure. In various embodiments, the antibody or protein configured to bind carbohydrate portions (e.g., lectin) is immobilized onto a bead substrate that is optionally a magnetic bead substrate. In one embodiment, the NSCLC is late-stage NSCLC. In one embodiment, the biological sample is a pre-treatment peripheral blood sample. In one embodiment, the one or more peptide structure includes a glycopeptide of a lung- specific protein, and the at least one peptide structure comprises three or more peptide structure identified in Table 3. In one embodiment, the at least one peptide structure comprises three or more amino acid sequence of SEQ ID NOs: 27-60 set forth in Table 3. In one embodiment, the at least one peptide structure comprises three or more amino acid sequence of SEQ ID NOs: 27-60 along with the associated glycan set forth in Table 3. In one embodiment, the at least one peptide structure comprises three or more amino acid sequence of SEQ ID NOs: 61-99 set forth in Table 3. In one embodiment, the at least one peptide structure comprises three or more amino acid sequence of SEQ ID NOs: 61-99 along with the associated glycan set forth in Table 3. In one embodiment, the at least one peptide structure comprises three or more peptide structures identified in Table 8. In one embodiment, the at least one peptide structure comprises three or more amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 set forth in Table 8. In one embodiment, the at least one peptide structure comprises three or more amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3.
[0158] In some embodiments, the method comprises assessing one or more clinical indicators of NSCLC. In some embodiments, the one or more clinical indicator of NSCLC comprises a new cough, a worsening cough, a persistent cough, a cough that produces blood (e.g., hemoptysis), shortness of breath, a persistent chest infection (e.g., bronchitis, pneumonia), wheezing, chest pain, voice hoarseness, headache, facial swelling, body swelling, upper body pain, weakness of the hand, a droopy eyelid, and blurred vision. In some embodiments, the method includes assessing a patient for unexplained weight loss, tiredness, fatigue, persistent or worsening bone pain, and persistent or worsening joint pain. In some embodiments, the method includes assessing the clinical indicators described herein and any combination thereof in diagnosing a patient having NSCLC. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of statesassociated with NSCLC based upon one or more peptides structure provided in Table 3 and / or Table 8. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more glycopeptides provided in Table 3 and / or Table 8. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC, based upon one or more peptide structure provided in Table 3 or Table 8, and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC, based upon one or more peptide structure provided in Table 3 or Table 8, and administering a treatment for NSCLC based upon the classification.
[0159] In some embodiments, the method comprises assessing one or more risk factors selected from the group comprising tobacco smoking, environmental and occupational carcinogen exposure (e.g., tar, soot, arsenic, chromium, and nickel, asbestos, radiation, radon), a family history of lung cancer, poor diet, and limited physical activity. In certain embodiments, the method comprises assessing one or more risk factors of NSCLC, wherein the risk factor of NSCLC is selected from the group consisting of tobacco smoking, a family history of lung cancer, environmental and occupational carcinogen exposure, poor diet, and limited physical activity. In certain embodiments, the individual is determined have a healthy state, in which a healthy state may include the absence of NSCLC. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of statesassociated with NSCLC based upon one or more peptide structures provided in Table 3 or Table 8. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99 In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58. In some embodiments, the diagnosis is based upon the presence and / or amount of each of the peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machinelearning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM- MS. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC, based upon one or more peptide structure provided in Table 3 or Table 8, and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC, based upon one or more peptide structure provided in Table 3 or Table 8, and administering a treatment for NSCLC based upon the classification.
[0160] FIG. 5 illustrates a flow diagram 500 of a method for classifying a biological sample obtained from a subject with respect to a plurality of states associated with NSCLC, in accordance with the presently disclosed embodiments. The flow diagram 500 may beperformed utilizing one or more processing devices (e.g., computing platform 302 as discussed above with respect to FIG. 3) that may include hardware (e.g., a general purpose processor, a graphic processing unit (GPU), an application- specific integrated circuit (ASIC), a system-on-chip (SoC), a microcontroller, a field-programmable gate array (FPGA), a central processing unit (CPU), an application processor (AP), a visual processing unit (VPU), a neural processing unit (NPU), a neural decision processor (NDP), a deep learning processor (DLP), a tensor processing unit (TPU), neuromorphic processing unit (NPU), or any other processing device(s) that may be suitable for processing various medical profile data and making one or more decisions based thereon), software (e.g., instructions running / executing on one or more processors), firmware (e.g., microcode), or some combination thereof.
[0161] The flow diagram 500 may begin at block 502 with one or more processing devices (e.g., computing platform 302) receiving peptide structure data corresponding to a set of glycoproteins in the biological sample. The flow diagram 500 may then continue at block 504 with one or more processing devices (e.g., computing platform 302) inputting quantification data identified from the peptide structure data for a set of peptide structures into a machine-learning model trained to identify a disease indicator based on the quantification data, wherein the set of peptide structures comprises at least one peptide structure identified from a plurality of peptide structures in Table 3 and / or Table 8. The flow diagram 500 may then continue at block 506 with one or more processing devices (e.g., computing platform 302) identifying, by the machine-learning model, the disease indicator. The flow diagram 500 may then conclude at block 508 with one or more processing devices (e.g., computing platform 302) classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the identified disease indicator.
[0162] FIG. 6 illustrates a flow diagram 600 of a method for detecting the presence of one of a plurality of states associated with NSCLC in a subject, in accordance with the presently disclosed embodiments. The flow diagram 600 may be performed utilizing one or more processing devices (e.g., computing platform 302 as discussed above with respect to FIG. 3) that may include hardware (e.g., a general purpose processor, a graphic processing unit (GPU), an application- specific integrated circuit (ASIC), a system-on-chip (SoC), a microcontroller, a field-programmable gate array (FPGA), a central processing unit (CPU), an application processor (AP), a visual processing unit (VPU), a neural processing unit (NPU), a neural decision processor (NDP), a deep learning processor (DLP), a tensorprocessing unit (TPU), neuromorphic processing unit (NPU), or any other processing device(s) that may be suitable for processing various medical profile data and making one or more decisions based thereon), software (e.g., instructions running / executing on one or more processors), firmware (e.g., microcode), or some combination thereof.
[0163] The flow diagram 600 may begin at block 602 with one or more processing devices (e.g., computing platform 302) receiving peptide structure data corresponding to a set of glycoproteins in a biological sample obtained from a subject, wherein the peptide structure data comprises at least one peptide structure from Table 3 and / or Table 8. The flow diagram 600 may then continue at block 604 with one or more processing devices (e.g., computing platform 302) inputting quantification data identified from the peptide structure data for a set of peptide structures into a machine-learning model trained to identify a disease indicator based on the quantification data, wherein the set of peptide structures comprises at least one peptide structure identified from a plurality of peptide structures in Table 3 and / or Table 8. The flow diagram 600 may then conclude at block 606 with one or more processing devices (e.g., computing platform 302) detecting the presence of a corresponding state of the plurality of states associated with NSCLC in response to a determination that the identified disease indicator falls within a selected range associated with the corresponding state.
[0164] FIG. 7 illustrates a flow diagram 700 of a method determining one or more of a plurality of treatment regimens for treating NSCLC in a subject, in accordance with the presently disclosed embodiments. The flow diagram 700 may be performed utilizing one or more processing devices (e.g., computing platform 302 as discussed above with respect to FIG. 3) that may include hardware (e.g., a general purpose processor, a graphic processing unit (GPU), an application- specific integrated circuit (ASIC), a system-on-chip (SoC), a microcontroller, a field-programmable gate array (FPGA), a central processing unit (CPU), an application processor (AP), a visual processing unit (VPU), a neural processing unit (NPU), a neural decision processor (NDP), a deep learning processor (DLP), a tensor processing unit (TPU), neuromorphic processing unit (NPU), or any other processing device(s) that may be suitable for processing various medical profile data and making one or more decisions based thereon), software (e.g., instructions running / executing on one or more processors), firmware (e.g., microcode), or some combination thereof.
[0165] The flow diagram 700 may begin at block 702 with one or more processing devices (e.g., computing platform 302) receiving peptide structure data corresponding to a set of glycoproteins in the biological sample. The flow diagram 700 may then continue at block 704 with one or more processing devices (e.g., computing platform 302) inputting quantification data identified from the peptide structure data for a set of the peptide structures into a machine-learning model trained to identify a disease indicator for NSCLC based on the quantification data, wherein the set of peptide structure data comprises at least one peptide structure identified from a plurality of peptide structures in Table 3 and / or Table 8. The flow diagram 700 may then continue at block 706 with one or more processing devices (e.g., computing platform 302) identifying, by the machine-learning model, the disease indicator. The flow diagram 700 may then conclude at block 708 with one or more processing devices (e.g., computing platform 302) determining at least one of a plurality of treatment regimens for treating NSCLC based upon the identified disease indicator.B. Training a Model
[0166] FIG. 8 illustrates a flow diagram 800 of a method for training a model to diagnose a subject with one of a plurality of states associated with NSCLC, in accordance with the presently disclosed embodiments. The flow diagram 800 may be performed utilizing one or more processing devices (e.g., computing platform 302 as discussed above with respect to FIG. 3) that may include hardware (e.g., a general purpose processor, a graphic processing unit (GPU), an application- specific integrated circuit (ASIC), a system-on-chip (SoC), a microcontroller, a field-programmable gate array (FPGA), a central processing unit (CPU), an application processor (AP), a visual processing unit (VPU), a neural processing unit (NPU), a neural decision processor (NDP), a deep learning processor (DLP), a tensor processing unit (TPU), neuromorphic processing unit (NPU), or any other processing device(s) that may be suitable for processing various medical profile data and making one or more decisions based thereon), software (e.g., instructions running / executing on one or more processors), firmware (e.g., microcode), or some combination thereof.
[0167] The flow diagram 800 may begin at block 802 with one or more processing devices (e.g., computing platform 302) receiving quantification data for a panel of peptide structures for a plurality of subjects diagnosed with the plurality of states associated with the NSCLC, wherein the quantification data comprises a plurality of peptide structure profiles forthe plurality of subjects and identifies a corresponding state of the plurality of states for each peptide structure profile of the plurality of peptide structure profiles. The flow diagram 800 may then conclude at block 804 with one or more processing devices (e.g., computing platform 302) training a machine-learning model to determine a state of the plurality of states a biological sample from the subject corresponds based on the quantification data.
[0168] In some embodiments, the machine-learning model includes a regularized regression model (e.g., LASSO regression model) trained and evaluated, for example, over one or more iterations using a first set of peptide structure coefficients set forth in Table 7 for all stages of NSCLC (stages 1-4), using a second set of peptide structure coefficients set forth in Table 7 for early-stage NSCLC (stages 1-2), and using a third set of peptide structure coefficients set forth in Table 7 for late-stage NSCLC (stages 3-4).C. Combined Diagnosis and Treatment
[0169] FIG. 9 illustrates a flow diagram 9000 of a method for treating NSCLC in a subject, in accordance with the presently disclosed embodiments. The flow diagram 900 may be performed utilizing one or more processing devices (e.g., computing platform 302 as discussed above with respect to FIG. 3) that may include hardware (e.g., a general purpose processor, a graphic processing unit (GPU), an application- specific integrated circuit (ASIC), a system-on-chip (SoC), a microcontroller, a field-programmable gate array (FPGA), a central processing unit (CPU), an application processor (AP), a visual processing unit (VPU), a neural processing unit (NPU), a neural decision processor (NDP), a deep learning processor (DLP), a tensor processing unit (TPU), neuromorphic processing unit (NPU), or any other processing device(s) that may be suitable for processing various medical profile data and making one or more decisions based thereon), software (e.g., instructions running / executing on one or more processors), firmware (e.g., microcode), or some combination thereof.
[0170] The flow diagram 900 may begin at block 902 with one or more processing devices (e.g., computing platform 302) receiving peptide structure data corresponding to a set of glycoproteins in the biological sample. The flow diagram 900 may then continue at block 904 with one or more processing devices (e.g., computing platform 302) inputting quantification data identified from the peptide structure data for a set of the peptide structures into a machine-learning model trained to identify a disease indicator based on thequantification data, wherein the set of peptide structure data comprises at least one peptide structure identified from a plurality of peptide structures in Table 3 and / or Table 8. The flow diagram 900 may then continue at block 906 with one or more processing devices (e.g., computing platform 302) identifying, by the machine-learning model, the disease indicator. The flow diagram 900 may then continue at block 908 with one or more processing devices (e.g., computing platform 302) determining a classification for NSCLC based upon the identified disease indicator. The flow diagram 900 may then conclude at block 910 with one or more processing devices (e.g., computing platform 302) determining a treatment to treat NSCLC based upon the classification.
[0171] FIG. 10 illustrates a flow diagram 1000 of a method for diagnosing an individual with NSCLC, in accordance with the presently disclosed embodiments. The flow diagram 1000 may be performed utilizing one or more processing devices (e.g., computing platform 302 as discussed above with respect to FIG. 3) that may include hardware (e.g., a general purpose processor, a graphic processing unit (GPU), an application-specific integrated circuit (ASIC), a system-on-chip (SoC), a microcontroller, a field-programmable gate array (FPGA), a central processing unit (CPU), an application processor (AP), a visual processing unit (VPU), a neural processing unit (NPU), a neural decision processor (NDP), a deep learning processor (DLP), a tensor processing unit (TPU), neuromorphic processing unit (NPU), or any other processing device(s) that may be suitable for processing various medical profile data and making one or more decisions based thereon), software (e.g., instructions running / executing on one or more processors), firmware (e.g., microcode), or some combination thereof.
[0172] The flow diagram 1000 may begin at block 1002 with one or more processing devices (e.g., computing platform 302) detecting the presence or amount of at least one structure structures from Table 3 and / or Table 8. The flow diagram 1000 may begin at block 1002 with one or more processing devices (e.g., computing platform 302) detecting the presence or amount of at least one peptide structure structures from Table 8. The flow diagram 1000 may then continue at block 1004 with one or more processing devices (e.g., computing platform 302) inputting a quantification of the detected at least one peptide structure into a machine-learning model trained to generate a class label. The flow diagram 1000 may then continue at block 1006 with one or more processing devices (e.g., computing platform 302) determining if the class label is above or below a threshold for a classification.The flow diagram 1000 may then continue at block 1008 with one or more processing devices (e.g., computing platform 302) identifying a diagnostic classification for a patient based on whether the class label is above or below a threshold for the classification. The flow diagram 1000 may then conclude at block 1010 with one or more processing devices (e.g., computing platform 302) diagnosing the patient as having NSCLC based on the diagnostic classification.VI. Methods of Treatment
[0173] In some embodiments, provided herein are methods of treating NSCLC based upon the presence, amount, and / or relative amount of one or more biomarkers provided herein. In some embodiments, the method comprises treating NSCLC based upon the presence, amount, and / or relative amount of one or more peptide structure from Table 3. In some embodiments, the method comprises treating NSCLC based upon the presence, amount, and / or relative amount of one or more peptide structures from Table 8. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structure provided herein and selecting a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structure provided in Table 3 and / or Table 8 and administering a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structure provided herein and administering a treatment for NSCLC based upon the classification. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structure provided in Table 3 and / or Table 8 and administering a treatment for NSCLC based upon the classification. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides to identify a disease indicator, detecting the presence of a corresponding state associated with NSCLC in response that the disease indicator falls within a selected range, and diagnosing NSCLC. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3. In some embodiments, the peptide structure data comprises one or more peptidestructure provided in Table 8. In some embodiments, the method further comprises administering an effective amount of a therapy for NSCLC. In some embodiments, the method further comprises selecting a particular therapy based upon the disease indicator.
[0174] In some embodiments, provided herein is a method of treating NSCLC comprising detecting the presence and / or amount of at least one peptide structure from Table 3 and selecting a NSCLC therapy. In some embodiments, method of treating NSCLC further comprises administering an effective amount of a NSCLC therapy to the individual based upon the presence and / or amount of at least one peptide structure from Table 3. In some embodiments, the diagnosis and / or treatment is based upon the presence and / or amount of at least two, at least three, at least four, at least five, at least 10, at least 15, at least 20, or at least 25 peptide structures from Table 3. In some embodiments, the method of treating NSCLC comprises detecting the presence (or absence) or amount of at least one peptide structure from Table 8 and selecting a NSCLC therapy. In some embodiments, the method of treating NSCLC comprises detecting the presence (or absence) or amount of at least one peptide structure from Table 8 and administering an effective amount of a NSCLC therapy to the individual. In some embodiments, the method further comprises selecting a therapy based upon the presence, and / or amount of the at least one peptide structure from Table 8. In some embodiments, the diagnosis and / or treatment is based upon the presence and / or amount of at least two, at least three, at least four, at least five, at least 10, or at least 15 peptide structures from Table 8.A. Treatments for NSCLC
[0175] In some embodiments, the method comprises selecting a therapy to treat NSCLC. In some embodiments, the method of selecting a therapy for NSCLC comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method of selecting a therapy comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the therapy is selected based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, at least ten, or at least 15 peptide structures from Table 3 and / or Table 8. In some embodiments, the therapy is selected based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, atleast ten, or at least 15 peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the therapy is selected on the basis of the stage of NSCLC. In some embodiments, the therapy is selected on the basis of one or more NSCLC risk factor in combination with the presence, absence, and or amount of one or more peptides or glycopeptides provided herein. In some embodiments, the therapy for NSCLC is selected from the group comprising a surgery, a chemotherapeutic therapy, a patient-specific therapy, a targeted immunotherapy, a radiation procedure, a radiofrequency ablation (RFA) procedure, or a combination thereof. In some embodiments, the surgery comprises the removal of one or more parts of the lung. In some embodiments, the platinum- coordinating compound comprises one of cisplatin, carboplatin, and nedaplatin. In some embodiments, the chemotherapeutic comprises docetaxel, paclitaxel, albumin-bound paclitaxel, vinorelbine, gemcitabine, irinotecan, pemetrexed, tegafur / gimeracil / oteracil, etoposide, or a combination thereof. In some embodiments, the chemotherapeutic therapy is a platinum-doublet regimen. In some embodiments, the chemotherapeutic therapy is a platinum-triple regimen. In some embodiments, the targeted immunotherapy comprises one or more antibody directed towards an immune system checkpoint protein including but not limited to PD-1, PD-L1, and CTLA-4. In some embodiments, the therapy for NSCLC comprises a combination of one or more antibody that targets PD-1, PD-L1, and CTLA-4. In some embodiments, the targeted therapy comprises one or more patient- specific therapy agent selected based on patient- specific changes in tumor cell gene expression including but not limited to changes in KRAS, EGFR, ALK, ROS1, BRAF, RET, MET, and NTRK genes. In some embodiments, the patient-specific therapy is an inhibitor of an oncogene. In some embodiments, the patient-specific therapy is an inhibitor of one or more of KRAS, EGFR, ALK, ROS1, BRAF, MEK, RET, MET, and NTRK. In some embodiments, the radiation procedure comprises the use of high-energy rays or particles to treat NSCLC. In some embodiments, the brachytherapy comprises the placement of radioactive material in or adjacent to the tumor in the airway (e.g., bronchial tubes).
[0176] In some embodiments, the method comprises administering a therapy to treat NSCLC. In some embodiments, the method of administering a therapy comprises inputting quantification data identified from peptide structure data for a set of peptides and / orglycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method of administering a therapy comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the therapy is administered based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, at least ten, or at least 15 peptide structures from Table 3 and / or Table 8. In some embodiments, the therapy is administered based upon presence and / or amount of at least one, at least two, at least three, at least four, at least five, at least ten, or at least 15 peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the therapy is administered on the basis of the stage of NSCLC. In some embodiments, the therapy is administered on the basis of one or more NSCLC risk factor in combination with the presence, absence, and or amount of one or more peptides or glycopeptides provided herein. In some embodiments, the therapy administered for NSCLC is selected from the group comprising a surgery, a chemotherapeutic therapy, a patient-specific therapy, a targeted immunotherapy, a radiation procedure, a radiofrequency ablation (RFA) procedure, or a combination thereof. In some embodiments, the surgery comprises the removal of one or more parts of the lung. In some embodiments, the platinum-coordinating compound comprises one of cisplatin, carboplatin, and nedaplatin. In some embodiments, the chemotherapeutic comprises docetaxel, paclitaxel, albumin-bound paclitaxel, vinorelbine, gemcitabine, irinotecan, pemetrexed, tegafur / gimeracil / oteracil, etoposide, or a combination thereof. In some embodiments, the chemotherapeutic therapy is a platinum-doublet regimen. In some embodiments, the chemotherapeutic therapy is a platinum-triple regimen. In some embodiments, the targeted immunotherapy comprises one or more antibody directed towards an immune system checkpoint protein including but not limited to PD-1, PD-L1, and CTLA- 4. In some embodiments, the therapy for NSCLC comprises a combination of one or more antibody that targets PD-1, PD-L1, and CTLA-4. In some embodiments, the patient- specific therapy comprises one or more patient- specific therapy agent selected based on patientspecific changes in tumor cell gene expression including but not limited to changes in KRAS, EGFR, ALK, ROS1, BRAF, RET, MET, and NTRK genes. In some embodiments, the patient-specific therapy is an inhibitor of an oncogene. In some embodiments, the patientspecific therapy is an inhibitor of one or more of KRAS, EGFR, ALK, ROS1, BRAF, MEK,RET, MET, and NTRK. In some embodiments, the radiation procedure comprises the use of high-energy rays or particles to treat NSCLC. In some embodiments, the brachytherapy comprises the placement of radioactive material in or adjacent to the tumor in the airway (e.g., bronchial tubes).
[0177] In some embodiments, the method comprises administering a therapy to treat NSCLC. In some embodiments, the method of administering a therapy comprises inputting quantification data identified from peptide structure data for a set of peptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the set of peptide structures comprising one or more, two or more, three or more, four or more, five or more, 10 or more, 15 or more, or each of the peptides and / or glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-60. In some embodiments, the method comprises inputting quantification data from peptide structure data into a machinelearning model, wherein the machine-learning model comprises a model trained using one or more of a first set of peptide structure coefficients set forth in Table 7 for all stages of NSCLC (stages 1-4), a second set of peptide structure coefficients set forth in Table 7 for early-stage NSCLC (stages 1-2), and a third set of peptide structure coefficients set forth in Table 7 for late-stage NSCLC (stages 3-4). In some embodiments, the model is trained using any one of the set of peptide structure coefficients described herein to identify a disease indicator, identifying by the machine-learning model, the disease indicator, and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the disease indicator. In some embodiments, the disease indicator comprises one or more scores that indicate a probability that the subject falls within one of the states associated with NSCLC (e.g., having NSCLC or not having NSCLC). In some embodiments, the model trained using any one of the first set, the second set, or the third set of peptide structure coefficients is a logical regression model. In some embodiments, the model trained using any one of the first set, the second set, or the third set of peptide structure coefficients comprises a LASSO regression model. In some embodiments, the peptide structures are detected using LC-MS. In some embodiments, the LC-MS comprises LC-MS / MS and LC-MS / MS running in a MRM mode. In some embodiments, the method comprises selecting a treatment for NSCLC based upon the classification described herein. In some embodiments, the method comprises administering a treatment for NSCLC based upon the classification described herein. In some embodiments, the therapy for NSCLC is selected from the group comprisinga surgery, a chemotherapeutic therapy, a patient- specific therapy, a targeted immunotherapy, a radiation procedure, a radiofrequency ablation (RFA) procedure, or a combination thereof. In some embodiments, the surgery comprises the removal of one or more parts of the lung. In some embodiments, the therapy comprises a lobectomy, a bronchial sleeve resection, a wedge resection, or a pneumonectomy. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptides is determined by MRM-MS. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method further comprises selecting a particular therapy described herein based upon the disease indicator and / or classification. In some embodiments, the method further comprises administering a particular therapy described herein based upon the disease indicator and / or classification.
[0178] In some embodiments, the chemotherapeutic therapy comprises a platinum- coordinating compound, a chemotherapeutic, or a combination thereof. In some embodiments, the platinum-coordinating compound comprises one of cisplatin (CDDP), carboplatin (CBDCA), and nedaplatin (CDGP). In some embodiments, the chemotherapeutic comprises docetaxel (Taxotere, DTX), paclitaxel (Taxol, PTX), albumin-bound paclitaxel (nab- paclitaxel, Abraxane), vinorelbine (Navelbine,VNR), gemcitabine (Gemzar, GEM), irinotecan (CPT-11), pemetrexed (Alimta, PEM), tegafur / gimeracil / oteracil (SI), etoposide (VP- 16), or a combination thereof. In some embodiments, the chemotherapeutic therapy is a platinum-doublet regimen. In some embodiments, the chemotherapeutic therapy is aplatinum-triple regimen. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptides is determined by MRM-MS. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method further comprises selecting a particular therapy described herein based upon the disease indicator and / or classification. In some embodiments, the method further comprises administering a particular therapy described herein based upon the disease indicator and / or classification.
[0179] In some embodiments, the targeted immunotherapy comprises one or more antibody directed towards an immune system checkpoint protein including but not limited to PD-1, PD-L1, and CTLA-4. In some embodiments, the antibody targeting PD-1 comprises nivolumab (Opdivo), pembrolizumab (Keytruda), and cemiplimab (Libtayo). In some embodiments, the antibody targeting PD-L1 comprises atezolizumab (Tecentriq), durvalumab (Imfinzi), and avelumab (Bavencio). In some embodiments, the antibody targeting CTLA-4 comprises ipilimumab (Yervoy). In some embodiments, the therapy for NSCLC comprises acombination of one or more antibody that targets PD-1, PD-L1, and CTLA-4. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more glycopeptides provided in Table 3 or Table 8In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method further comprises selecting a particular therapy described herein based upon the disease indicator and / or classification. In some embodiments, the method further comprises administering a particular therapy described herein based upon the disease indicator and / or classification.
[0180] In some embodiments, the therapy comprises one or more patient- specific therapy agent selected based on patient- specific changes in tumor cell gene expression including but not limited to changes in KRAS, EGFR, ALK, ROS1, BRAF, RET, MET, and NTRK genes. In some embodiments, the patient-specific therapy is an inhibitor of an oncogene. In some embodiments, the patient-specific therapy is an inhibitor of one or more of KRAS, EGFR,ALK, R0S1, BRAF, MEK, RET, MET, and NTRK. In some embodiments, the patientspecific therapy comprises one or more of sotorasib (Lumakras), erlotinib (Tarceva), afatinib (Gilotrif), gefitinib (Iressa), osimertinib (Tagrisso), dacomitinib (Vizimpro), amivantamab (Rybrevant), nobocertinib (Exkivity), necitumumab (Portrazza), crizotinib (Xalkori), ceritinib (Zykadia), alectinib (Alecensa), brigatinib (Alunbrig), lorlatinib (Lorbrena), entrectinib (Rozlytrek), dabrafenib (Tafinlar), trametinib (Mekinist), selpercatinib (Retevmo), pralsetinib (Gavreto), capmatinib (Tabrecta), tepotinib (Tepmetko), larotrectinib (Vitrakvi), and combinations thereof. In some embodiments the patient-specific therapy comprises an angiogenesis inhibitor. In some embodiments, the angiogenesis inhibitor comprises one of bevacizumab (Avastin, BEV) and ramucirumab (Cyramza, RAM). In some embodiments, the therapy for NSCLC comprises a combination of one or more patient- specific therapy agents. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more glycopeptides provided in Table 3 or Table 8. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising theglycopeptides provided in Table 3 or Table 8. In some embodiments, the method further comprises selecting a particular therapy described herein based upon the disease indicator and / or classification. In some embodiments, the method further comprises administering a particular therapy described herein based upon the disease indicator and / or classification.
[0181] In some embodiments, the radiation procedure comprises the use of high-energy rays or particles to treat NSCLC. In some embodiments, the radiation procedure comprises external beam radiation therapy (EBRT) and internal radiation therapy (also referred to as brachytherapy). In some embodiments, the EBRT comprises one or more of stereotactic ablative radiotherapy (SABR), three-dimensional conformal radiation therapy (3D-CRT), intensity modulated radiation therapy (IMRT), volumetric modulated arc therapy (VMAT), and stereotactic radiosurgery (SRS). In some embodiments, the brachytherapy comprises the placement of radioactive material in or adjacent to the tumor in the airway (e.g., bronchial tubes). In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more glycopeptides provided in Table 3 or Table 8. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8.In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method further comprises selecting a particular therapy described herein based upon the disease indicator and / or classification. In some embodiments, the method further comprises administering a particular therapy described herein based upon the disease indicator and / or classification.
[0182] In some embodiments, the method involves monitoring of the individual for progression of NSCLC. In some embodiments, the method involving monitoring comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into a machine-learning model trained to identify a disease indicator. In some embodiments, the machine-learning model is a regularized regression model (e.g., LASSO regression model) trained using one or more of a first set, a second set, or a third set of peptide structure coefficients set forth in Table 7. In some embodiments, the first set, the second set, or the third set of peptide structure coefficients used to train the model comprise one or more of the amino acid sequence of SEQ ID NOs: 27-60 as set forth in Table 7. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the method involving monitoring comprises classifying a subsequent biological sample with respect to a plurality of states associated with NSCLC, based upon one or more peptide structures provided in Table 3 and / or Table 8, and selecting a treatment for NSCLC based upon the classification. In some embodiments, the monitoring comprises classifying a subsequent biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structures provided in Table 3 and / or Table 8 and administering a treatment for NSCLC based upon the classification. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the method involving monitoring comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides to identify a disease indicator, detecting the presence of a corresponding state associated with NSCLC in response that the disease indicator falls within a selected range, and diagnosing NSCLC. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the peptide structure datacomprises one or more glycopeptide structure provided in Table 3 and / or Table 8. In some embodiments, the method involving monitoring further comprises selecting a particular therapy based upon the disease indicator. In some embodiments, the method involving monitoring further comprises administering an effective amount of a therapy for NSCLC.B. Patient Populations
[0183] Provided herein is a method of diagnosis and treatment for an individual. Further provided herein is a method of diagnosis and treatment for an individual with one or more risk factors associated with NSCLC. In some embodiments, the method comprises measuring the amount / presence or absence of one or more peptides structures from Table 3 in an individual with one or more risk factors associated with NSCLC. In some embodiments, the method involves diagnosing an individual based upon presence and / or amount of one or more peptide structures from Table 3. In some embodiments, the method involves diagnosing an individual based upon presence and / or amount of one or more glycopeptides from Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more peptides and / or glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27- 99 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99 set forth in Table 3. In some embodiments, the diagnosis is based upon the presence and / or amount of one or more glycopeptides consisting of the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, the individual diagnosed with NSCLC is administered one or more NSCLC therapies described herein, based on the disease indicator determined by the diagnosis. In some embodiments, the individual is administered one or more NSCLC therapies described herein, based on the disease indicator determined by the diagnosis. In some embodiments, the individual confirmed to have NSCLC is treated based on the disease indicator determined by the diagnosis.
[0184] In some embodiments, the individual is diagnosed, wherein one or more peptide structures from Table 3 are detected and are distinct from a healthy control sample. In some embodiments, the individual is diagnosed, wherein one or more peptide structures comprising the amino acid sequence of SEQ ID NOs: 27-99 are detected and are distinct from a healthy control sample. In some embodiments, the individual is diagnosed, wherein one or moreglycopeptides comprising the amino acid sequence of SEQ ID NOs: 27-99 are detected and are distinct from a healthy control sample. In some embodiments, the amount of at least one peptide structure is none, or below a detection limit. In some embodiments, the amount of at least one glycopeptide structure is none, or below a detection limit. In some embodiments, the amount of at least one peptide structure from Table 3 is none, or below a detection limit. In some embodiments, the amount of at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 27-99 set forth in Table 3 is none, or below a detection limit. In some embodiments, the amount of at least one peptide structure is significantly lower than a control sample from a healthy individual. In some embodiments, the amount of at least one glycopeptide structure is significantly lower than a control sample from a healthy individual. In some embodiments, the amount of at least one peptide structure from Table 3 is significantly lower than a control sample from a healthy individual. In some embodiments, the amount of at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 27-99 set forth in Table 3 is significantly lower than a control sample from a healthy individual. In some embodiments, the amount of at least one peptide structure is significantly higher than a control sample from a healthy individual. In some embodiments, the amount of at least one glycopeptide structure is significantly higher than a control sample from a healthy individual. In some embodiments, the amount of at least one peptide structure from Table 3 is significantly higher than a control sample from a healthy individual. In some embodiments, the amount of at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 27-99 set forth in Table 3 is significantly higher than a control sample from a healthy individual. In some embodiments, the individual is diagnosed and treated according to the presence and / or amount of one or more peptide structures from Table 3. In some embodiments, the individual is diagnosed and treated according to the presence and / or amount of one or more peptide structures comprising the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3.
[0185] In some embodiments, the individual has NSCLC. In some embodiments, the individual has stage 0, stage I, stage II, stage III, or stage IV NSCLC. In some embodiments, the individual has early-stage NSCLC. In some embodiments, the individual has late-stage NSCLC or advanced NSCLC. In some embodiments, the individual has NSCLC that has not spread from the site of origination. In some embodiments, the individual has NSCLC that has spread locally to the surrounding tissue. In some embodiments, the individual has NSCLCthat has spread beyond the original tumor and / or the local tumor environment. In some embodiments, the individual has NSCLC that has spread to one or more organs beyond the lungs. In some embodiments, the individual has metastatic NSCLC. In some embodiments, the individual has NSCLC and has relapsed and / or progressed. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises classifying a biological sample with respect to a plurality of states associated with NSCLC based upon one or more glycopeptides provided in Table 3 or Table 8. In some embodiments, the method comprises inputting quantification data identified from peptide structure data for a set of peptides and / or glycopeptides into one or more machine-learning model trained to identify a disease indicator. In some embodiments, the method comprises classifying the sample as having NSCLC or not having NSCLC based upon the disease indicator. In some embodiments, the peptide structure data comprises one or more peptide structure provided in Table 3 and / or Table 8. In some embodiments, the presence, absence, and / or amount of one or more peptides and / or glycopeptide is determined by MRM-MS. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises selecting a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the peptide structures provided in Table 3 or Table 8. In some embodiments, the method comprises administering a particular therapy described herein based upon the presence, amount, and / or relative amount of one or more biomarkers comprising the glycopeptides provided in Table 3 or Table 8. In some embodiments, the method further comprises selecting a particular therapy described herein based upon the disease indicator and / or classification. In some embodiments, the method further comprises administering a particular therapy described herein based upon the disease indicator and / or classification.
[0186] In some embodiments, the individual has had prior lines of therapy for treating NSCLC. In some embodiments, the individual has had at least 1, at least 2, or at least 3 priorlines of therapy for treating NSCLC. In some embodiments, the individual has had no more than 1, no more than 2, or no more than 3 prior lines of therapy for treating NSCLC. In some embodiments, the individual has not had prior therapy for treating NSCLC.
[0187] In some embodiments, the individual has altered gene expression relevant for NSCLC treatment. In some embodiments, the individual has altered oncogene expression. In some embodiments, the individual has altered tumor cell gene expression. In some embodiments, the altered gene expression comprises altered gene expression of one or more of KRAS, EGFR, ALK, ROS1, BRAF, RET, MET, and NTRK. In some embodiments, the altered gene expression comprises altered gene expression of one or more immune system checkpoint proteins PD-1, PD-L1, and CTLA-4. In some embodiments, the individual having altered gene expression relevant for NSCLC treatment may benefit from a therapy comprising one or more antibody that targets PD-1, PD-L1, and CTLA-4, or a combination thereof.
[0188] In some embodiments, the individual is at risk of developing NSCLC. In some embodiments, the individual is positive for one or more risk factor that increases the chances of developing NSCLC. In some embodiments, the one or more risk factor is smoking, wherein the individual smokes at least one of cigarettes, cigars, pipes, and other tobaccobased products. In some embodiments, the individual is a smoker. In some embodiments, the individual smoked in the past and has quit smoking. In some embodiments, the individual is or has been a recreational smoker, wherein smoking occurs infrequently. In some embodiments, the individual is a non-smoker and has never smoked before. In some embodiments, the individual is exposed to secondhand smoke. In some embodiments, the individual is exposed to environmental or occupational carcinogens that increase the risk of developing NSCLC. In some embodiments, the environmental or occupational carcinogen comprises one or more of asbestos, arsenic, chromium, beryllium, nickel, soot, tar, indoor air pollution, and outdoor air pollution. In some embodiments, the individual is exposed to radiation that increase the risk of developing NSCLC. In some embodiments, the radiation comprises one or more of radiation therapy for the breast or chest, extensive CT scan imaging, and radon exposure. In some embodiments, the individual has a family history of lung cancer that increase the risk of developing NSCLC. In some embodiments, the individual is aging, wherein the individual is at least 40 years old. In some embodiments, theindividual has at least 1, at least 2, at least 3, at least 4, at least 5, or at least 6 risk factors for NSCLC.
[0189] In some embodiments, the individual is at least 18 years old. In some embodiments, the individual is at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, at least 60, at least 65, at least 70, at least 75, at least 80, at least 85, or at least 90 years old. In some embodiments, the individual is at least 65 years old.
[0190] In some embodiments, the individual is positive for one or more disease features of NSCLC. In some embodiments, the one or more disease features of NSCLC comprise a new cough, a worsening cough, a persistent cough, a cough that produces blood (e.g., hemoptysis), shortness of breath, a persistent chest infection (e.g., bronchitis, pneumonia), wheezing, chest pain, voice hoarseness, a headache, facial swelling, body swelling, upper body pain, weakness of the hand, a droopy eyelid, blurred vision, unexplained weight loss, tiredness, fatigue, persistent or worsening bone pain, and persistent or worsening joint pain. In some embodiments, the individual has at least 1, at least 2, at least 3, at least 4, at least 5, or at least 6 disease features of NSCLC. In some embodiments, the individual has any combination of disease features of NSCLC described herein.VII. Compositions and Kits
[0191] Provided herein are compositions comprising one or more peptide structures from Table 3. Provided herein are compositions comprising one or more glycopeptides from Table 3. In some embodiments, provided herein is a composition comprising two or more peptide structures from Table 3. In some embodiments, provided herein is a composition comprising three or more peptide structures from Table 3. In some embodiments, provided herein is a composition comprising four or more peptide structures from Table 3. In some embodiments, provided herein is a composition comprising five or more peptide structures from Table 3. In some embodiments, provided herein is a composition comprising 10 or more peptide structures from Table 3. In some embodiments, provided herein is a composition comprising 15 or more peptide structures from Table 3. In some embodiments, provided herein is a composition comprising 20 or more peptide structures from Table 3. In some embodiments, provided herein is a composition comprising 25 or more peptide structures from Table 3. In some embodiments, the composition is from a biological sample. In some embodiments, thecomposition comprises one or more purified peptide structures. In some embodiments, the composition comprises one or more purified glycopeptides. In some embodiments, the composition comprises enzymatically digested peptide and / or glycopeptide fragments, such as those in Table 3. In some embodiments, the composition comprises enzymatically digested glycopeptide fragments, such as those in Table 3. In some embodiments, the composition comprises at least one, at least two, at least three, at least four, at least five, at least 10, at least 15, at least 20, or at least 25 peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3.
[0192] In some embodiments, provided herein is a composition comprising one or more peptide structures from Table 8. In some embodiments, provided herein is a composition comprising one or more glycopeptides from Table 8. In some embodiments, provided herein is a composition comprising two or more peptide structures from Table 8. In some embodiments, provided herein is a composition comprising three or more peptide structures from Table 8. In some embodiments, provided herein is a composition comprising four or more peptide structures from Table 8. In some embodiments, provided herein is a composition comprising five or more peptide structures from Table 8. In some embodiments, provided herein is a composition comprising 10 or more peptide structures from Table 8. In some embodiments, provided herein is a composition comprising 15 or more peptide structures from Table 8. In some embodiments, the composition is from a biological sample. In some embodiments, the composition comprises one or more purified peptide structures. In some embodiments, the composition comprises one or more purified glycopeptides. In some embodiments, the composition comprises enzymatically digested peptide fragments, such as those in Table 8. In some embodiments, the composition comprises enzymatically digested glycopeptide fragments, such as those in Table 8. In some embodiments, the composition comprises at least one, at least two, at least three, at least four, at least five, at least 10, or at least 15 peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3.
[0193] In some embodiments, provided herein is a composition comprising at least one peptide and / or glycopeptide comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least two peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In someembodiments, provided herein is a composition comprising at least three peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least four peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least five peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least 10 peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least 15 peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising 20 peptides and / or glycopeptides comprising sequences set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising 25 peptides and / or glycopeptides comprising sequences set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3.
[0194] In some embodiments, provided herein are peptides and / or glycopeptides set forth in Table 3. In some embodiments, provided herein are glycopeptides set forth in Table 3. In some embodiments, provided herein are peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3.
[0195] In some embodiments, provided herein is a composition comprising at least one peptide and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least two peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least three peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least four peptides and / or glycopeptides comprising a sequence setforth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least five peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least 10 peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, provided herein is a composition comprising at least 15 peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3.
[0196] In some embodiments, provided herein are peptides and / or glycopeptides set forth in Table 8. In some embodiments, provided herein are glycopeptides set forth in Table 8. In some embodiments, provided herein are peptides and / or glycopeptides comprising a sequence set forth in SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3.
[0197] Also provided herein are kits for detecting NSCLC in an individual and kits for treating NSCLC in an individual. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure identified in Table 3. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure identified in Table 8. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure, wherein the agent is from a biological sample, is one or more purified peptide structures, and / or comprises enzymatically digested peptide and / or glycopeptide fragments. In some embodiments, the kit comprises instructions for detecting one or more biomarkers provided herein, wherein the biomarker is at least one peptide structure identified in Table 3. In some embodiments, the kit comprises instructions for detecting one or more biomarkers provided herein, wherein the biomarker is at least one glycopeptide identified in Table 3. In some embodiments, the kit comprises instructions fordetecting one or more biomarkers provided herein, wherein the biomarker is at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, the kit comprises instructions for detecting one or more biomarkers provided herein, wherein the biomarker is at least one peptide structure identified in Table 8. In some embodiments, the kit comprises instructions for detecting one or more biomarkers provided herein, wherein the biomarker is at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3. In some embodiments, the kit comprises reagents for detecting one or more biomarkers described herein. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure identified in Table 3 to carry out part or all of any one or more of the methods disclosed herein. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3 to carry out part or all of any one or more of the methods disclosed herein. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure identified in Table 8 to carry out part or all of any one or more of the methods disclosed herein. In some embodiments, the kit comprises at least one agent for quantifying at least one peptide structure comprising the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forth in Table 3 to carry out part or all of any one or more of the methods disclosed herein.
[0198] In some embodiments, the kit comprises at least one of a peptide standard, a glycopeptide standard, a buffer, and at least one or more peptide sequences to carry out part or all of any one or more of the methods disclosed herein. In some embodiments, the kit comprises at least one of a peptide standard, a glycopeptide standard, a buffer, and at least one or more glycopeptide sequences to carry out part or all of any one or more of the methods disclosed herein. In some embodiments, the one or more peptide and / or glycopeptide sequences are identified in Table 3. In some embodiments, the one or more peptide and / or glycopeptide sequences comprise the amino acid sequence of SEQ ID NOs: 27-99 along with the associated glycan set forth in Table 3. In some embodiments, the one or more peptide and / or glycopeptide sequences are identified in Table 8. In some embodiments, the one or more peptide and / or glycopeptide sequences comprise the amino acid sequence of SEQ ID NOs: 32, 34-37, 39, 42, 44, 47-48, 50-58 along with the associated glycan set forthin Table 3. In some embodiments, the kit comprises instructions for detecting NSCLC in an individual. In some embodiments, the kit comprises instructions for diagnosing NSCLC in an individual. In some embodiments, the kit comprises instructions for use according to any of the methods provided herein.
[0199] In some embodiments, the kit comprises instructions for selecting a treatment for the individual having NSCLC. In some embodiments, the kit comprises instructions for administering a treatment to the individual having NSCLC.
[0200] In some embodiments, the kit comprises reagents and instructions for selecting a treatment for an individual with NSCLC, wherein the treatment is one or more NSCLC treatments described herein. In some embodiments, the kit comprises reagents and instructions for administering the treatment to an individual with NSCLC comprising administering to the individual one or more NSCLC treatments described herein. In some embodiments, the kit comprises reagents and instructions for treating an individual diagnosed with NSCLC comprising selecting one or more NSCLC treatments described herein and / or administering to the individual one or more NSCLC treatments described herein. In some embodiments, the kit comprises instructions for use according to any of the methods provided herein.VIII. Enumerated Embodiments1. A method of classifying a biological sample obtained from a subject with respect to a plurality of states associated with non-small cell lung cancer (NSCLC), the method comprising receiving peptide structure data corresponding to a set of proteins in the biological sample; inputting quantification data identified from the peptide structure data for a set of peptide structures into a machine-learning model trained to identify a disease indicator based on the quantification data, wherein the set of peptide structures comprises at least one peptide structure identified from a plurality of peptide structures in Table 12B; identifying, by the machine-learning model, the disease indicator; andclassifying the biological sample with respect to a plurality of states associated with NSCLC based upon the identified disease indicator.2. A method of detecting the presence of non-small cell lung cancer (NSCLC) in a subject, the method comprising receiving peptide structure data corresponding to a set of proteins in a biological sample obtained from a subject, wherein the peptide structure data comprises at least one peptide structure from Table 12B; inputting quantification data identified from the peptide structure data for a set of peptide structures into a machine-learning model trained to identify a disease indicator based on the quantification data; and detecting the presence of NSCLC in response to a determination that the identified disease indicator falls within a selected range associated with NSCLC.3. The method of embodiment 1 or 2, wherein the set of proteins comprises one or more glycoproteins.4. The method of any one of embodiments 1-3, wherein the at least one peptide structure comprises a glycopeptide.5. The method of any one of embodiments 1-4, wherein the plurality of states comprises at least one of a NSCLC state or a healthy state.6. The method of any one of embodiments 1-5, wherein the machine-learning model comprises a regularized regression model.7. The method of any one of embodiments 1-6, wherein the regularized regression model comprises a least absolute shrinkage and selection operator (LASSO) regression model.8. The method of any one of embodiments 1-7, wherein the quantification data for a peptide structure of the set of peptide structures comprises at least one of an abundance, a relative abundance, a normalized abundance, or a differential abundance.9. The method of any one of embodiments 1-8, wherein the quantification data for a peptide structure of the set of peptide structures comprises at least one of a relative quantity, an adjusted quantity, a normalized quantity, a relative concentration, an adjusted concentration, or a normalized concentration.10. The method of any one of embodiments 1-9, wherein the quantification data is generated using a liquid chromatography -mass spectrometry (LC-MS) system.11. The method of any one of embodiments 1-10, wherein the peptide structure data is generated using multiple reaction monitoring mass spectrometry (MRM-MS).12. The method of any one of embodiments 1-11, wherein the machine-learning model was trained utilizing a portion of the quantification data corresponding to a set of peptide structures that is a subset of the panel of peptide structures to determine which state of the plurality of states the biological sample from the subject corresponds.13. The method of any one of embodiments 1-12, further comprising: performing a differential expression analysis using the quantification data for the plurality of subjects.14. The method of any one of embodiments 1-13, wherein the biological sample comprises at least one of blood, serum, or plasma.15. A method of diagnosing an individual with non-small cell lung cancer (NSCLC), comprising: detecting the presence or amount of at least one peptide structure structures from Table 12B; inputting a quantification of the detected at least one peptide structure into a machinelearning model trained to generate a class label; determining if the class label is above or below a threshold for a classification; identifying a diagnostic classification for the individual based on whether the class label is above or below a threshold for the classification; anddiagnosing the individual as having NSCLC based on the diagnostic classification.16. The method of any one of embodiments 1-15, wherein the quantification data is generated using a liquid chromatography -mass spectrometry (LC-MS) system.17. The method of any one of embodiments 1-16, wherein the peptide structure data is generated using multiple reaction monitoring mass spectrometry (MRM-MS).18. The method of any one of embodiments 1-17, wherein the amount of at least one peptide structure is none, or below a detection limit.19. The method of any one of embodiments 1-18, wherein the NSCLC is one of early- stage or late-stage NSCLC.20. The method of any one of embodiments 1-19, wherein the NSCLC is one of stage I NSCLC, stage II NSCLC, stage III NSCLC, or stage IV NSCLC.21. The method of any one of embodiments 1-20, wherein the at least one peptide structure comprises three or more peptide structures identified in Table 12B.22. The method of any one of embodiments 1-21, wherein the at least one peptide structure comprises at least one peptide comprising the sequence set forth in any one of SEQ ID NOs: 101-142.23. The method of any one of embodiments 1-22, wherein the subject has one or more risk factors associated with NSCLC.24. The method of any one of embodiments 1-23, wherein the individual is determined have a healthy state, wherein a healthy state comprises the absence of NSCLC.25. The method of any one of embodiments 1-24, further comprising generating a report that includes a diagnosis based on the corresponding state detected for the subject.26. The method of any one of embodiments 1-25, wherein the at least one peptide structure comprises at least one, at least two, at least three, at least five, at least 10, at least 15, at least 20, or at least 25 different peptides comprising the sequence set forth in any one of SEQ ID NOs: 101-142.27. The method of any one of embodiments 1-25, wherein the at least one peptide structure comprises a peptide sequence and a glycan structure, wherein the glycan structure is attached to a linking site position in the peptide sequence in accordance with Table 12B.28. The method of embodiment 27, wherein the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 12B, wherein the glycan structure comprises a symbol structure in accordance with the glycan structure GL number according to Table 12B, Table 17, and Table 18.29. The method of embodiment 27 or 28, wherein the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 12B, wherein the glycan structure comprises a composition in accordance with the glycan structure GL number, Table 12B, Table 17, and Table 18.30. The method of any one of embodiments 1-29, wherein a rightmost N-acetylgalactosamine of the glycan structure in Table 18 is attached to a linking site position in the peptide sequence in accordance with Table 12B, and wherein a bottommost N-acetylglucosamine of the glycan structure in Table 17 is attached to a linking site position in the peptide sequence in accordance with Table 12B.31. A composition comprising one or more peptide structures from Table 12B.32. A composition comprising one or more peptides comprising the sequence set forth in SEQ ID NOs: 101-142.33. The composition of embodiment 32, wherein the one or more peptides comprise one or more glycopeptides.34. The composition of embodiment 31, wherein the at least one peptide structure comprises a peptide sequence and a glycan structure, wherein the glycan structure is attached to a linking site position in the peptide sequence in accordance with Table 12B.35. The composition of embodiment 34, wherein the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 12B, wherein the glycan structure comprises a symbol structure in accordance with the glycan structure GL number according to Table 12B, Table 16, and Table 17.36. The composition of embodiment 34 or 35, wherein the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 12B, wherein the glycan structure comprises a composition in accordance with the glycan structure GL number, Table 12B, Table 16, and Table 17.37. The composition of any one of embodiments 34-36, wherein a rightmost N-acetylgalactosamine of the glycan structure in Table 18 is attached to a linking site position in the peptide sequence in accordance with Table 12B, and wherein a bottommost N-acetylglucosamine of the glycan structure in Table 17 is attached to a linking site position in the peptide sequence in accordance with Table 12B.IX. Examples
[0201] The invention will be more fully understood by reference to the following examples. They should not, however, be construed as limiting the scope of the invention. It is understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this application and scope of the appended claims.Example 1. Sample Preparation and MS Data Production for Training a Diagnostic Model
[0202] A schematic for the overall workflow for sample preparation and analysis is given in FIG. 11, in which the data interpretation process as illustrated by Step 6 may include performing one or more machine-learning model based prediction and / or classification tasksutilizing a regularized regression model (e.g., LASSO regression model) as described herein. In this workflow, 694 glycopeptide (GP) and non-glycosylated peptide transitions were interrogated, which were derived from 74 serum proteins in pre-treatment peripheral blood samples from a cohort of 194 healthy individuals and a cohort of 316 individuals with NSCLC, as shown in Table 1.
[0203] Individuals were assessed prior to sample analysis, using clinical criteria for the NSCLC diagnosis and staging. Lung biopsy samples were subject to a histopathological assessment and clinical imaging techniques were used to stage individuals having NSCLC. The stage of the NSCLC was assigned using the TNM system that is based on the size and extent of the main tumor (T), the spread to nearby lymph nodes (N), and the spread (metastasis) to distant sites (M).Table 1. Summary of patient age and sexstage.
[0204] Pooled human serum for assay normalization and calibration purposes, dithiothreitol (DTT), and iodoacetamide (IAA) were purchased from Millipore Sigma (St. Louis, MO). Sequencing grade trypsin was purchased from Promega (Madison, WI). Acetonitrile (LC-MS grade) was purchased from Honeywell (Muskegon, MI). All other reagents used were procured from Millipore Sigma, VWR, and Fisher Scientific.
[0205] Prior to analysis, plasma samples were reduced with DTT in a water bath at 60°C for 50 min, then alkylated with IAA followed by digestion with trypsin in a water bath at 37°C for 18 hours. To quench the digestion, formic acid was added to each sample after incubation to a final concentration of 1% (v / v). A pool of 18 stable isotope-labeled synthetic peptides matching the sequence of 18 endogenous peptide targets were included at a knownconcentration for the purpose of determining absolute endogenous protein concentrations in the samples.
[0206] Digested plasma samples were injected into an Agilent 6495B triple quadrupole mass spectrometer equipped with an Agilent 1290 Infinity ultra-high-pressure (UHP)-LC system and an Agilent ZORBAX Eclipse Plus C18 column (2.1 mmxl50 mm i.d., 1.8 pm particle size). Separation of the peptides and glycopeptides was performed using a 49-min binary gradient. The aqueous mobile phase A was 3% acetonitrile, 0.1% formic acid in water (v / v), and the organic mobile phase B was 90% acetonitrile, 0.1% formic acid in water (v / v). The flow rate was set at 0.5 mL / min. Electrospray ionization (ESI) was used as the ionization source and was operated in positive ion mode. The triple quadrupole MS was operated in dynamic multiple reaction monitoring (dMRM) mode. Samples were injected in a randomized fashion with regard to underlying phenotype, and reference pooled serum digests were injected interspersed with study samples, at every 10thsample position throughout the run. The R library ‘caret’ was used for building machine-learning models, and all other statistical analyses used base R functions.
[0207] Concentration of peptides and glycopeptides were calculated by the following formula:
[0208] Wherein the internal standard (ISTD) corresponds to the peptide of interest, optionally as a stable isotope internal standard (e.g., heavy labeled lysine). An example ISTD could be GWVTDGFSSLK* where the terminal lysine is a heavy stable isotope labeled lysine for the unlabeled peptide with sequence - GWVTDGFSSLK. glycopeptide raw abundance glycopeptide site occupancy = — - - - - - glycoprotein raw abundance
[0209] Wherein the glycoprotein raw abundance is the raw abundance sum of all of the glycopeptides from the same glycoprotein.[gly copeptide]approximate= glycopeptide site occupancy x [peptide]giycopeptide
[0210] Wherein the [peptide] is the peptide concentration of the quantification peptide from the same glycoprotein.Example 2. Analysis of Peptide Structure Data
[0211] A MRM-MS analysis was performed on pre-treatment peripheral blood samples from healthy individuals and individuals having NSCLC, wherein digested and processed samples containing one or more of peptide and glycopeptide were subjected to LC-MS / MS running in a MRM mode to quantify the presence, amount, and / or structure of the one or more peptides and glycopeptide. Concentrations of glycopeptides and peptides were calculated as described in Example 1. Provided in Table 3 are the identified peptide sequences and glycan moieties that were determined to have significantly different concentrations for peptides and glycopeptides when comparing the healthy cohort vs the NSCLC cohort. In Tables 4A and 4B, the glycan structures referenced in Table 3 are shown. The proteins and glycoproteins associated with these peptides and glycopeptides, respectively, are summarized in Table 2. In Table 5, the MRM-MS and liquid chromatography (LC) parameters for the peptide and glycopeptide identification are summarized.Table 2. Proteins and glycoproteins associated with NSCLC control, all stages of NSCLC, early-stage NSCLC, and late-stage NSCLC
[0212] Table 3 includes the Peptide Structure (PS) Name (e.g., A2GL_186_5402), which is a reference code for the protein name (e.g., A2GL), followed by the glycan linking site position in the protein (e.g., the number 186 that is preceded by an underscore and represents a sequential amino acid position in protein A2GL), and followed by the glycan structure GL number (e.g., the number 5402 that is preceded by an underscore and represents a glycan composition Hex(5)HexNAc(4)Fuc(0)NeuAc(2)). The Peptide Structure (PS) Name of Table 3 contains a prefix that represents an abbreviation (that may include a combination of letters and numbers) for a protein abbreviation that corresponds to the Protein Abbreviation of Table 2. The term Glycan Linking Site Pos. in Protein Sequence is a number that refers to the sequential position of an amino acid of the corresponding protein in which a glycan isattached. For the Glycan Linking Site Pos. in Protein Sequence, the amino acid position of the peptide sequence is defined by the sequentially numbered order of amino acids based on the Uniprot ID of the corresponding protein for the peptide sequence. The term Glycan Linking Site Pos. in Peptide Sequence is a number that refers to the sequential position of an amino acid of the corresponding peptide in which a glycan is attached. For the Glycan Linking Site Pos. in peptide Sequence, the amino acid position of the peptide sequence is defined by the sequentially numbered order of amino acids for the peptide sequence. The term Glycan Structure GL No. is a number that corresponds to a symbol structure and a composition of the glycan as indicated in Tables 4A and 4B.Table 3. Details of peptides and glycopeptides with statistically significant changing abundances in healthy and NSCLC samples
[0213] In some embodiments, the term AGP12 for SEQ ID NOs: 67-71 represent that the glycopeptide is a fragment of either AGP1 or AGP2. In some embodiments, the term IGA12 for SEQ ID NOs: 46, 86-89 represent that the glycopeptide is a fragment of either IGA1 or IGA2. For the SEQ ID NO:79 in Table 3, the identity of the glycopeptide is one of two possibilities that have the same monoisotopic mass. In the first possibility, the glycan having the Glycan GL NO 6513 is attached to the peptide with a Glycan linking site position of 5 inthe peptide sequence. In the second possibility, the glycan having the Glycan GL NO 6502 is attached to the peptide with a Glycan linking site position of 9 in the peptide sequence.
[0214] In Table 3 (and Table 12B), if the first number subsequent to the first underscore in the Peptide Structure (PS) NAME is inconsistent with the Glycan Linking Site Pos. in Protein Sequence column, then the Glycan Linking Site Pos. in Protein Sequence column should be used for identification of the peptide. In Table 3 (and Table 12B), if the second number subsequent to the second underscore in the Peptide Structure (PS) NAME is inconsistent with the Glycan Structure GL NO column, then the Glycan Structure GL NO column should be used for identification of the glycan portion of the glycopeptide. If the Peptide Structure (PS) NAME does not contain any numbers, then the peptide is nonglycosylated. In some instances of the Peptide Structure (PS) NAME, subsequent to the prefix, there is a number noted with the notation MC that indicates that there was a missed cleavage at position in the peptide sequence as noted by the number. In some instances of the Peptide Structure (PS) NAME, there is a suffix NH3LOSS to indicate a loss of a NH3 group.
[0215] Tables 4A and 4B (and also Table 17 and Table 18) illustrate the symbol structure and composition of detected glycan moieties that correspond to glycopeptides of Table 3 (and Table 12B) based on the Glycan GL NO. The term Symbol Structure illustrates a geometric linking structure of the carbohydrates where the bottommost carbohydrate such as N-acetylglucosamine is bound to the designated amino acid for an N-linked glycan and the rightmost carbohydrate such as N-acetylgalactosamine is bound to the designated amino acid for an O-linked glycan. It should be noted that the Glycan Structure GL NOs. 1102 and 2201 are O-linked glycans that are in Table 4 A and that N-linked glycans are in Table 4B. For reference, N-linked glycans have a glycan attached to the amino acid asparagine and O- linked glycans have a glycan attached to either a serine or a threonine.
[0216] The identity of the various monosaccharides is illustrated by the Legend section located at the end of Table 4B (or Table 18). The abbreviations of the Legend are Glc that represents glucose and is indicated by a dark circle, Gal that represents galactose and is indicated by an open circle, Man that represents mannose and is indicated by a circle with intermediate grey shading, Fuc that represents fucose and is indicated by a dark triangle, Neu5Ac that represents N-acetylneuraminic acid and is indicated by a dark diamond, GlcNAc that represents N-acetylglucosamine and is indicated by a dark square, GalNAc thatrepresents N-acetylgalactosamine and is indicated by an open square, and ManNAc that represents N-acetylmannosamine and is indicated by a square with intermediate grey shading.
[0217] The term Composition refers to the number of various classes of carbohydrates that make up the glycan. The quantity for each class of carbohydrate is depicted as a number in parenthesis to the right of an abbreviation that corresponds to the class of the carbohydrate. The abbreviations for these clasess are Hex, HexNAc, Fuc, and NeuAc that respectively correspond to hexose, N-acetylhexosamine, fucose, and N- acetylneuraminic acid. It should be noted that hexose sugars include glucose, galactose, and mannose; and N- acetylhexosamine sugars includes N-acetylglucosamine, N-acetylgalactosamine, and N- acetylmannosamine. In various embodiments, the terms Neu5Ac, NeuAc, and N- acetylneuraminic acid may be referred to as sialic acid.
[0218] Referring back to Tables 4A and 4B (or Table 17 and Table 18), for some entries, there are two symbol structures provided for one Glycan Structure GL NO such as, for example, Glycan Structure GL NO 4500. Thus, the identify of a peptide that references a Glycan Structure GL NO that has two symbol structures could one of two possibilities based on the MRM of the LC-MS analysis. In some instances, a bracket symbol is used as part of the Symbol Structure to indicate that the precise bonding linkage is not exactly known, but that the linking line segment is attached to one of the plurality of adjacent carbohydrates immediately adjacent to the bracketTable 4A. Glycan structure GL NO, symbol structure, and composition of detected glycan moieties for O-linked glycansTable 4B. Glycan structure GL NO, symbol structure, and composition of detected glycan moieties for N-linked glycansLegend for Tables 4A and 4BTable 5. MRM-MS and liquid chromatography (LC) parameters for peptide structures associated with healthy control and NSCLC populations
[0219] Table 5 (or Table 13 below) shows various parameters associated with the identification of the peptide and glycopeptides using LC and MRM-MS. The retention time (RT) represents the amount of time in minutes for the peptide elute from the chromatography column. The collision energy represents the energy applied to the peptide for creating fragments (i.e., product ions) such as, for example, in the 2ndquadrupole of the triple quadrupole MS. The first precursor m / z represents a ratio value associated with an ionized form having a first precursor charge for the peptide or glycopeptide. Similarly, the second precursor m / z represents a ratio value associated with an ionized form having a second precursor charge for the peptide or glycopeptide. The first precursor ion is associated with a first product ion having a m / z ratio that was formed from a collision and the second precursor ion is associated with a second product ion having a m / z ratio that was formed from a collision. Under certain circumstances, the first precursor and the second precursor may be the same, but the associated first and second product m / z ratios are different.
[0220] To demonstrate the statistical significance in the peptide structure concentration difference between the healthy control and NSCLC populations (stages 1-4 population, stages 1-2 population, or stages 3-4 population), the fold changes, p-values, false discovery rates (FDR), and area under the curved (AUC) using one marker in the model are provided in Tables 6 A and 6B. Fold-changes for individual peptides and glycopeptides, were calculated on normalized abundances of NSCLC stages 1-4 samples vs control (Table 6A). NSCLC stages 1-2 samples vs control (Table 6B), and NSCLC stages 3-4 vs control samples (Table 6B). For example, the fold change can be based the concentration of a biomarker from one of the NSCLC cohorts divided by the concentration of the same biomarker from the control cohort. In general, fold changes that are much higher or lower than unity are associated with a difference between the two cohorts. False discovery rate was calculated using the Benjamini-Hochberg method. Age- and sex-adjusted differential expression analysis for 596 normalized biomarkers were performed to evaluate statistically significant differential abundances using an FDR-adjusted q-value of 0.05 as a cutoff. Repeated five-fold cross-validated LASSO-regularized logistic regression was performed to create a multivariable classifier that predicts whether a serum sample belongs to the healthy or NSCLC cohort.Table 6A. Differential expression analysis for healthy control and NSCLC (stages 1-4) sample setsTable 6B. Differential expression analysis for healthy control and NSCLC (stages 1-2 or stages 3-4) sample setsExample 3. Training and Validation of Diagnostic Models
[0221] The quantified concentrations of various peptide structures identified in Table 3 (e.g., SEQ. ID NOs: 27-99) across the entire sample set were used to train a regularized regression model (e.g., LASSO regression model) to generate a disease indicator for a subject. The disease indicator was generated as a score (e.g., a probability score), wherein the range in which the score falls enables diagnosis or classification as a healthy state or a NSCLC state. The same markers were used to train a regularized regression model (e.g., LASSO regression model). Coefficients for the regularized regression model (e.g., LASSO regression model) are provided in Table 7.Table 7. Coefficients for each marker used in the regularized regression model (e.g., LASSO regression model)
[0222] Using the values of Table 7, a probability can be determined by summing together the product of the concentration of each biomarker in the sample and the respective coefficient (of one column) and then adding the summation and the intercept to yield the logit of a probability score. For example, the logit of the probability, to which the inverse logit function can be applied, is equal to:
[0223] With respect to the above equations and Table 7, coefficients that were N / A could be construed as zero. Using a FDR < 0.05 in comparing NSCLC and healthy control samples, 432 biomarkers with significant abundance differences were identified. Using 70% of the complete cohort (balanced by case / control membership, NSCLC stage, sex, and age quartile) as a training set, a total of 375 glycopeptide and non-glycosylated peptide biomarker features were selected that remained differentially expressed at FDR-adjusted q-value < 0.05 as input into a LASSO-regularized multivariable classifier. This resulted in a 19-biomarker model (Table 8) exhibiting an accuracy of 94.8% (96.9% sensitivity, 91.2% specificity) and AUC of 0.989 in predicting NSCLC for all stages (i.e., stages 1-4). This classifier was validated in an independent test set including the remaining 30% of subjects, yielding an accuracy of 94.5% (95.5% sensitivity, 93.0% specificity) and AUC of 0.975. Sensitivity in the test set was 100% / 96% / 99% / 96% / 94% / 100%, for NSCLC stages 0, 1, 2, 3, 4, and “unknown”, respectively. It should be noted that stage 0 is a cohort where a carcinoma is in situ and is considered the earliest stage at which lung cancer can be detected. The stage “unknown” refers to samples that have NSCLC, but the information regarding stage is unknown.
[0224] FIG. 12 is a model performance plot 1200 illustrating the predictive performance of a regularized regression model (e.g., LASSO regression model) trained and evaluated on the training data set and test data set splits for all stages of NSCLC (stages 1-4). FIG. 14 is a model performance plot 1400 illustrating the predictive performance of the regularized regression model (e.g., LASSO regression model) trained and evaluated on the training data set and test data set splits for early-stage NSCLC (stages 1-2). FIG. 16 is a model performance plot 1600 illustrating the predictive performance of the regularized regression model (e.g., LASSO regression model) trained and evaluated on the training data set and test data set splits for late-stage NSCLC (stages 3-4).
[0225] FIG. 13 illustrates a receiver-operating-characteristic (ROC) curve / area under curve (AUC) plot 1300, depicting that the regularized regression model (e.g., LASSO regression model) trained and evaluated on the training data set and test data set splits for all stages of NSCLC (stages 1-4) exhibited an accuracy of 94.8% and 94.5% on the training dataset and test data set, respectively, and AUC of 0.98 in distinguishing between healthy and NSCLC individuals. FIG. 15 illustrates a receiver-operating-characteristic (ROC) curve / area under curve (AUC) plot 1500, depicting that the regularized regression model (e.g., LASSO regression model) trained and evaluated on the training data set and test data set splits for early-stage NSCLC (stages 1-2) exhibited an accuracy of 91.7% and 92.7% on the training data set and test data set, respectively, and AUC of 0.98 in distinguishing between healthy and NSCLC individuals. FIG. 17 illustrates a receiver-operating-characteristic (ROC) curve / area under curve (AUC) plot 1700, depicting that the regularized regression model (e.g., LASSO regression model) trained and evaluated on the training data set and test data set splits for late-stage NSCLC (stages 3-4) exhibited an accuracy of 94.8% and 86.2% on the training data set and test data set, respectively, and AUC of 0.98 in distinguishing between healthy and NSCLC individuals.Table 8. Peptide and glycopeptide structures where coefficients were calculated for a model predicting probability of having NSCLC based on control vs stages 1-4
[0226] This result demonstrates that the identified peptide structures in Table 3 and a trained model using the peptide structures can be used to diagnose NSCLC.X. Additional Embodiments for Detection of Non-Small Cell Lung Cancer
[0227] Background: Non-small cell lung cancer (NSCLC) remains a leading cause of cancer death since most patients present at later stages. Abnormal protein glycosylation provides an avenue for innovative NSCLC biomarker discovery. This study aimed to identify circulating glycoproteomic biomarkers for NSCLC detection.
[0228] Methods: Utilizing a novel glycoproteomic profiling platform that combines liquid-chromatography / mass- spectrometry and artificial-intelligence-powered data processing, we assessed glycopeptide (GP) and non-glycosylated peptide quantification transitions in pre-treatment peripheral blood serum samples from NSCLC, COPD, and healthy individuals.
[0229] Results: We detected 354, 195, and 340 GPs / peptides with statistically significant abundance differences (FDR<0.05), when comparing NSCLC (N=372) with healthy controls (N=494), COPD (N=148), and NSCLC-negative (COPD + healthy controls), respectively. Among these detected GPs / peptides, 50 GP biomarkers were significantly differentially expressed between NSCLC and both COPD and healthy controls separately, but not between healthy controls and COPD samples. Multivariate machine learning classifiers were developed. When we modeled on NSCLC and healthy controls only, the classifier yielded an AUC of 0.956 (sensitivity 91.6%, specificity 90.3%). We then developed a classifier trained on healthy controls and COPD samples in the NSCLC-negative set providing greater patient heterogeneity. This classifier achieved 85.5% specificity to detect healthy controls, 84.2% sensitivity to detect all stages of NSCLC, and an AUC of 0.875 in the held-out test set. It detected both early (stage I- II) and late (stage III- IV) stages of NSCLC with sensitivities of 81.8% and 87.2%, respectively.
[0230] Conclusion: Using a novel platform, we identified serum glycoproteomic profiles which can be applied to the development of non-invasive NSCLC screening tools.INTRODUCTION
[0231] Lung cancer is the leading cause of cancer deaths in the world, with about 2.2 million new cases and 1.8 million deaths worldwide. Non-small cell lung cancer (NSCLC) is the most common type of lung cancer, accounting for about 85% of all lung cancer cases. The 5-year survival rate for lung cancer is about 22% for all stages combined. However, the survival rate is much higher for people whose cancer is diagnosed at an early stage before it has metastasized; the five-year survival rate for people with stage I lung cancer is about 60%, plummeting to about 6% for those with stage IV. These data underscore the critical importance of screening and early diagnosis in preventing the morbidity and mortality associated with late-stage diagnosis and treatment of lung cancer.
[0232] Although low dose computed tomography (LDCT) has been studied and is recommended by the US Preventive Services Task Force for early detection of lung cancer, its effectiveness remains uncertain despite the significant reduction in lung cancer-related mortality demonstrated by the National Lung Screening Trial (NLST) study. One disadvantage highlighted by the NLST is the high false positive rate of 96%, as LDCT can identify common incidental pulmonary findings. Moreover, clinicians must consider the potential harms of serial LDCT scans, which can increase the risk of radiation-induced cancer. The adoption of LDCT has been limited, with estimated screening rates of less than 4% due to the expense and logistical challenges of implementing LDCT screening. Additionally, LDCT screening targets a selected population with heavy smoking history, whereas screening and early diagnosis beyond this population remain underdeveloped.
[0233] Liquid biopsies have been touted as an effective screening modality. This involves the non-invasive analysis of circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), and exosomes in a patient's blood which not only can capture real-time changes in tumor biology but also has the potential to overcome the limitations of tissue sampling in cases where a biopsy may be risky or difficult to obtain. The drawback is the enormous heterogeneity of alterations seen at the ctDNA level as well as the overwhelming prominence of these types of changes in people with late- stage cancer. This underscores the unmet need for non-invasive screening alternatives.
[0234] Recently, the use of serum glycoproteomics as a measure of host response to disease has gained prominence as a screening, diagnostic, and prognostic tool. Protein glycosylation is the most abundant and complex form of post-translational protein modification and represents a novel paradigm for biomarker discovery in the context of lung cancer early detection. For example, Kondo et al. has demonstrated the difference in N- glycosylation status between NSCLC and small-cell lung cancer. Other studies have suggested the potential of serum N-glycomics signatures to distinguish subjects diagnosed with lung cancer from controls. In blinded pre-diagnostic serum samples, glycan variables were identified to have significant discriminatory power between NSCLC cases and controls and may have predictive value for assessing risk of lung cancer.
[0235] In this study, the potential of differential site-specific protein glycosylation as biomarkers in NSCLC was evaluated using a novel platform that combines liquid chromatography / mass spectrometry with a proprietary artificial-intelligence-based data processing engine.MATERIALS AND METHODSPatient samples
[0236] Serum samples were obtained from NSCLC patients, patients with chronic obstructive pulmonary disease (COPD) with no current malignancies prior to blood draw, and healthy individuals who are presumed to be NSCLC-negative. NSCLC samples were sourced from Biobank Graz of the Medical University of Graz (Graz, Austria) and the commercial biorepository iSpecimen (Lexington, MA, USA). COPD samples were sourced from Indivumed (Hamburg, Germany), and a portion of the healthy samples were provided by Precision for Medicine (Frederick, MD, USA). A prospective InterVenn- sponsored study (South San Francisco, CA, USA) provided the remaining portion of the NSCLC-negative healthy samples (NCT05445570). Samples were collected in accordance with relevant applicable guidelines and regulations for human subjects' protection as outlined in the Declaration of Helsinki. All protocols were approved under respective Institutional Review Boards and Ethics Committee (20223899, 20221935, 32-601). All subjects utilized in this study provided a written informed consent prior to collection of samples. Collected clinical data included age at blood draw, sex, and when applicable, histopathological data and clinical staging.Reagents and preanalytical sample preparation
[0237] Pooled human serum (MilliporeSigma, St. Louis, MO) was used for quality control, assay normalization, and calibration purposes. Dithiothreitol (DTT) and iodoacetamide (IAA) were purchased from MilliporeSigma (St. Louise, MO). LC-MS grade trypsin, acetonitrile, and formic acid were purchased from Thermo Scientific (Waltham, MA, USA), and stable isotope-labeled peptides were purchased from Vivitide (Gardner, MA, USA). Serum samples were treated with dithiothreitol (DTT) and iodoacetamide (IAA) to reduce disulfide bonds and alkylate sulfhydryls, respectively, followed by digestion with trypsin at 37°C for 18 hours. The digestion was quenched by adding formic acid to each sample to a final concentration of 1% (v / v) followed by the addition of a cocktail of stable isotope-labeled peptide internal standards at known concentrations.Liquid chromatography-mass spectrometry (LC-MS)
[0238] Digested serum samples were separated using a Waters (Milford, MA, USA) ACQUITY Peptide HSS T3 column (2.1 mm internal diameter x 150 mm length, 1.8 pm particle size) on an Agilent 1290 Infinity UHPLC system. The mobile phase A consisted of 0.1% formic acid in water (v / v), and the mobile phase B of 0.1% formic acid in acetonitrile (v / v), with the flow rate set at 0.5 mL / minute. After UHPLC separation, the peptides and glycopeptides (GPs) were ionized by electrospray ionization in positive ion mode on an Agilent 6495C triple quadrupole mass spectrometer before they were quantified in a targeted manner by dynamic multiple reaction monitoring (dMRM). PeakBoundaryNetl6, an in- house-developed spectrogram feature recognition and integration software based on recurrent neural networks, was used to integrate chromatogram peaks and to obtain molecular abundance quantification for each peptide and glycopeptide. The 694 MRM transitions, which represent 75 high-abundance (concentrations of gj 10 pg / ml or gj 200nM) serum glycoproteins, comprised 532 GPs which were normalized by comparing them with the abundance of their respective endogenous non-glycosylated peptide fragment, each fragment representing each of the 75 proteins from which the GPs monitored were derived. Samples were injected in a randomized fashion with respect to underlying phenotype, and reference pooled serum digests were injected interspersed with study samples. This workflow is visualized in Figure 18.Data Analysis
[0239] Quality control procedures were performed to filter out analytes with high coefficients of variation, high variability across between-run patient sample replicates, and extreme outliers due to technical artifacts. In total, 487 analytes (423 GPs and 64 nonglycosylated peptides) were deemed acceptable for further analysis and classifier development. To perform data processing and develop machine learning classifiers, the R packages ‘stats’, ‘dplyr’, ‘glmnet’, and ‘caret’ were used. Figures were generated using ‘gplots’, ‘ggplot2’, and ‘ggbeeswarm’ R packages.
[0240] Normalized abundances of GPs and peptides, corrected for between-run and within-run drift, were quantified in all patient samples described above. Raw abundances of non-glycosylated peptides were normalized by using spiked-in heavy-isotope-labeled internal standards (ISTDs) with known peptide concentrations. Raw abundance of glycopeptides was normalized by dividing by the abundance of their respective endogenous non-glycosylated peptide fragment to yield a relative abundance measure. The 423 relative abundance-normalized GPs and 64 ISTD concentration-normalized peptides were all log-transformed such that their expression profiles more closely approximate the Gaussian distribution. A scatterplot of the top two principal components from the dimension reduction method principal component analysis colored by phenotype was rendered to demonstrate visual differentiation of assay-wide expression levels between phenotype groups.
[0241] Age- and sex-adjusted differential expression analysis was performed via linear regression with clinical phenotype serving as the primary binary independent variable for each biomarker. We evaluated statistically significant differential abundances between several pairs of phenotypes. To correct for multiple comparisons, the Benjamini-Hochberg false discovery rate (FDR) was used, and 0.05 was selected as the cutoff to determine statistical significance. Hierarchically clustered heatmaps of normalized expression filtered by conditions such as differential expression statistical significance in particular contrasts and grouped by phenotype were rendered to reveal differentiated expression patterns between phenotypes in biomarkers of interest. Sets of differentially expressed biomarkers identified in the full sample cohort were not considered in classifier development, as it is critical to avoid the leakage of information from validation and test sets into the training set.
[0242] Repeated five-fold cross-validated elastic net-regularized logistic regression was performed to develop multivariate classifiers mentioned below that predict whether a serum sample belongs to the NSCLC-positive or NSCLC-negative cohort. In the first model, we compared NSCLC with healthy controls (as NSCLC-negative). In the second model, we compared NSCLC with the combination of healthy controls and COPD samples (as NSCLC- negative). The full sample set was split by phenotype, NSCLC stage, and source into training, validation, and test sets comprising 60%, 15%, and 25% of the sample set, respectively, such that all three groups contained approximately identical proportions of each (Table 11). Classifier coefficients were fit in the training set, and hyperparameters were tuned to prevent overfitting in the validation set. When the classifier is trained on both healthy controls and COPD, the probability threshold was selected to index on 90% sensitivity to detect NSCLC in the training set. The probability threshold chosen in the training set in all multivariate classifiers was also applied to predicted probabilities in the validation and test sets such that patients achieving a probability greater than the defined threshold were classified as NSCLC- positive, and NSCLC-negative otherwise. Classifiers were applied to the held-out test set only after all classifying determinations, such as hyperparameter tuning described above,were complete. ROC curves for each set as well as performance metrics stratified by phenotype and disease stage were generated.Table 11. Demographic covariates in the full cohort stratified by sample set. Counts are followed by the appropriate column-wise percentage, while continuous variables are summarized by means and standard deviations.RESULTSPatients and healthy controls characterization.
[0243] We interrogated 694 GP and non-glycosylated peptide quantification transitions in pre-treatment peripheral blood serum samples from several cohorts (Table 9). The NSCLC cohort included 372 individuals (242 from iSpecimen and 130 from Biobank Graz) with NSCLC (158 females, 214 males; mean age of 65.5 years (SD 9.6); stage I-IV sample sizes, respectively: 134 / 81 / 83 / 70, 4 with missing stage) at time of blood draw. There were 494 healthy controls at time of blood draw (235 females, 259 males; mean age of 59.1 years (SD 10)). 148 patients were those with COPD who had no malignancies prior to blood draw (68 males, 80 females; mean age of 64.2 years (SD 11.6)) and were sourced from Indivumed. In our study there were 642 individuals without NSCLC.Table 9. Demographic covariates in the full cohort stratified by phenotype. Counts are followed by the appropriate column-wise percentage, while continuous variables are summarized by means and standard deviation.Differential glycoproteomic profiles in NSCLC, COPD, and healthy controls.
[0244] There were 340 differentially expressed GP and non-glycosylated peptide biomarkers with statistically significant differences in normalized abundance at FDR<0.05 when comparing the two contrasting phenotype groups of interest, healthy controls and COPD combined (n=642) and NSCLC (n=372). To understand the contrasts in a more targeted way, we compared healthy controls (n=494) and NSCLC (n=372) as well as COPD (n=148) and NSCLC (n=372), yielding 354 and 195 biomarkers with statistically significant abundance differences, respectively. Lastly, to understand the glycoproteomic biomarkers that may be differentially expressed due to the underlying COPD disease process, including generalized acute or chronic inflammation, we compared healthy controls (n=494) and COPD (n=148), yielding 309 statistically significant biomarkers. These inherent differences between phenotype groups are reflected visually in the PCA scatterplot (Figure 19A).
[0245] There were overlaps of differentially expressed biomarkers that were of interest (Figure 19B). Fifty biomarkers were significantly differentially expressed between healthy controls and NSCLC samples, as well as between COPD and NSCLC samples, but not between healthy controls and COPD samples; this set is likely to be NSCLC -specific and not as a result of the variations related to non-malignant processes like an inflammatory response among COPD and healthy individuals (Figure 20). Modest separation between the phenotypes, especially those with respect to NSCLC, suggested that differential glycoproteomic profiles could be exploited for the development of a multivariate classifier.Multivariate classifier performance
[0246] To first show that a blood glycoproteomic profile -based multivariate classifier can be developed to identify NSCLC, we fit a classifier trained on only healthy controls and NSCLC samples. The resulting classifier exhibited 92.4% sensitivity and 91.6% specificity with an AUC of 0.974 in the training set. After tuning, the classifier was applied to the held- out test set, yielding a sensitivity of 91.6% and a specificity of 90.3% with an AUC of 0.956. Sensitivity to detect early and advanced stage NSCLC in the test set was 89.1% and 94.9%, respectively. However, this classifier predicts a vast majority of COPD patients as NSCLC (Figure 23). Despite this classifier strongly differentiating healthy controls and individuals with NSCLC, there is motivation to develop a classifier more reflective of the complicated population encountered in clinical practice.
[0247] Therefore, we developed a classifier trained on not only NSCLC and healthy controls, but on COPD samples as well; the inclusion of COPD samples in the NSCLC- negative samples set provides greater patient heterogeneity in the NSCLC-negative cohort and thus closer emulates a realistic screening population. A new classifier was trained under this new paradigm and achieved 87.8% specificity to detect healthy controls when indexing on approximately 90% sensitivity to detect NSCLC of all stages in the training set (Table 10). The classifier retained 42 biomarkers, 37 of which were relative abundance-normalized glycopeptides and five of which were ISTD concentration-normalized non-glycosylated peptides (Table 12A). After tuning, the classifier was applied to the held-out test set and achieved 85.5% specificity to detect healthy controls and 84.2% sensitivity to detect NSCLC (stage I: 79.4%, stage II: 85.7%, stage III: 90.5%, stage IV: 83.3%). The specificity to detect all controls, healthy and COPD, in the test set was 76.4%. ROC curves show a well-trained classifier with little to no overfitting, with the test set AUC achieving 0.875 (Figure 21A). Distributions of predicted probabilities in each phenotype and stage group show an increase in probabilities across groups, with the COPD group behaving as an intermediate class between healthy controls and those with NSCLC, and minimal difference in performance between early stage and advanced stage NSCLC, with sensitivities of 81.8% and 87.2%, respectively (Table 10 and Figure 21B). Distributions of predicted probabilities across sample sets also show little to no overfitting. The patient samples in the held-out test set are displayed in a heatmap filtered to the biomarkers retained in the classifier, demonstrating again the differentiation of expression between phenotypes (Figure 22).Table 10. Performance of the classifier in the training, validation, and held-out test sample sets stratified by phenotype and NSCLC stage.Table 12A. Differentially expressed biomarkers between the groups healthy controls / COPD and NSCLC that are in the multivariate model.Characterization of glycoproteins retained in the multivariate classifier.
[0248] A total of 42 markers were retained in the above multivariate classifier (Table 12A). These were largely liver-derived glycoproteins (33 markers, 78.5%) with a smaller but significant subset being immunoglobulin-related (9 markers, 21.4%). Protein- wise, 5 of the markers were derivatives of immunoglobulin G (IgG), 3 were each of apolipoprotein E (APOE) and kininogen-1 (KNG1), 2 each of alpha-2-HS -glycoprotein (FETUA), complement factor H (CFAH), and immunoglobulin M (IgM) (Figure 24A). 5 (-12%) of the markers were peptides while the remaining 37 represent relative abundances of specific glycoforms. 3 of these were non-glycosylated glycoforms and the remaining GPs weremostly N-glycosylated (32, 86.5%) while 2 had O-glycans (5.4%). Of the N-glycosylated forms, 23 (62.1%) bore bi-antennary glycans while 9 had either tri- or tetra-antennary N- glycans (Figure 24B). Of all N-glycosylated forms, 13 were fucosylated while 19 did not have a fucose.DISCUSSION:
[0249] Improvements in lung cancer screening are imperative to reduce the currently observed morbidity and mortality. In recent years, increasing incidence and prevalence of stage I NSCLC in patients younger than 65 years of age has been observed. Early detection with LDCT screening improves a patient’s prognosis and reduces mortality associated with lung cancer; yet, screening rates remain low at only 5-7%. Current lung cancer screening guidelines exclude non-smokers from lung screening trials and programs; however, studies suggest that rates of lung cancer in non-smokers are increasing overall. Given restricted access and limited utility of current lung cancer screening methods, there is a need for an accurate, non-invasive test for NSCLC detection.
[0250] With our high resolution and high throughput platform, combining LC-MS and machine learning data processing, we demonstrated distinct differential glycoproteomic profiles between NSCLC, COPD and healthy controls, offering unique and novel insights into aberrant protein glycosylation during carcinogenesis. Among the observed markers, 50 biomarkers were significantly differentially expressed between NSCLC and both COPD and healthy controls separately, but not between healthy controls and COPD samples. The separation of glycoproteomic profiles among the phenotypes, especially those with respect to NSCLC, made it possible to develop a multivariate classifier for NSCLC based on aberrant glycoproteomic profiles. We observed the characterization of these glycoforms that were retained in the multivariate machine learning based analysis. We note two major sources of biomarkers in our panel, the first being immune-derived such as glycoforms of IgG, IgM, and immunoglobulin J (IgJ), and the second are liver-derived glycoproteins such as APOE, KNG1, and FETUA. The effects of differential glycosylation of IgG have long been known, galactosylation, sialylation, and fucosylation all can affect function. These changes in glycosylation lead to inhibitory or anti-inflammatory effects, to activation of complement, and in turn to promotion of antibody-dependent cellular cytotoxicity. It remains to be seen which epitopes are recognized by the IgGs we detect on our panel, and how their differential glycosylation affects their function. In our later model, we noticed a significant contributionof glycoform biomarkers from the three liver-derived glycoproteins mentioned earlier. APOE has been implicated to be involved in promoting lung adenocarcinoma proliferation and migration, but little is known about how differentially glycosylated versions of this protein affect these neoplastic events. APOE has also recently been classified as a checkpoint inhibitor of the complement system by coupling with component Iq (Clq). This glycosylation-dependent mechanism may also be relevant during lung cancer pathogenesis. Both KNG1 and FETUA have been studied as biomarkers of lung cancer at the proteome level but ours is the first study that looks at specific glycoforms of these glycoproteins.
[0251] Based on differential glycoproteomic profiles among NSCLC, COPD and healthy controls, we developed multivariate classifiers. The samples included in this study were sourced from independent cohorts in Austria, the United States, and Germany, and include both retrospective and prospective collections. To ensure a broadly applicable signal in NSCLC, non-cancerous COPD samples were included as controls. This heterogeneity in source and cohorts, as well as the rigorous split of these samples into training and validation loops, with a held-out test set, lead to the robustness of our classifier to identify NSCLC. The diverse cohorts allowed us to develop and compare more than one classifier. The primary classifier (i.e. second model) we have developed has a high sensitivity to detect NSCLC (84.2%), particularly in early stages (81.8%), while maintaining high specificity for detecting healthy controls (85.5%). By training the classifier on COPD as well as healthy controls, glycopeptide and peptide biomarkers implicated in the classifier are more likely to be NSCLC-specific rather than those dysregulated by lung inflammation and other non- malignant processes alone. The alternative classifier (i.e., first model) that was developed without training on COPD had a much lower specificity to classify COPD as NSCLC- negative (10.8%) when compared to the primary classifier (45.9%) (see Figure 21B for primary classifier and Figure 23 for alternative classifier). This was attributed to the magnitude of differentiation between NSCLC-negative controls and NSCLC samples on which the alternative classifier was trained. This differentiation was greater than what is seen when COPD samples are trained on as part of the NSCLC-negative sample set (Figure 21B, Figure 23). While the alternative classifier has a specificity and sensitivity greater than 90% to detect healthy controls and NSCLC respectively, most of the glycopeptides and peptides implicated are presumed to be detecting generalized non-malignant conditions like inflammation instead of a NSCLC-specific signal. Nevertheless, both classifiers showed potential strengths and limitations. Further investigation, including improved modeling andcombining the test with clinical evaluations (e.g., history or clinical signs of COPD), are underway.
[0252] Development of biomarkers starts with the initial discovery of potential biomarkers, proceeds with a thorough assessment of their classification accuracy, and culminates in evaluating the biomarkers’ impact on clinical outcomes. Our development program therefore has started with the curation of high-quality specimen collections to assess the biomarkers’ ability to discriminate between NSCLC and non-cancer states like COPD. Thus, this study carries the limitations one can expect during this discovery phase. This includes the limitation that the cohorts used were analyzed in concert based on availability alone, and the proportions between phenotypes therein do not necessarily mirror real- world observations. In particular, the relative disproportional nature of our control cohort is likely quite different compared to a true screening cohort for NSCLC commonly observed in practice. Ideally, a multivariate classifier that screens for a condition as insidious as NSCLC would be trained on many more inflammatory and benign conditions to increase clinical utility. This is typically reserved for the next phase of biomarker development. We recognize that even slight differences in the processing and handling of a biospecimen may result in variation in the reproducibility and reliability of the study. Given the exploratory nature of the study, it is not uncommon practice to have specimens that originate from multiple sources. Nevertheless, it is promising to observe an exploratory classifier with a high sensitivity and specificity in a heterogeneous cohort. An additional limitation of our study is that we were not able to factor in relevant covariates such as smoking status into our inclusion criteria or univariate analysis. Evidence suggests that smoking affects changes in N- glycosylation. Although we have observed a set of differentially expressed biomarkers, we may have identified a slightly different set if we had been able to adjust for smoking status and other clinical variables in the analysis. Finally, as this was an exploratory study, we utilized the same biospecimen sources for our training, validation, and testing cohorts. Our findings would benefit from external validation studies using truly independent cohorts collected from lung screening studies.
[0253] The field of lung cancer screening and early diagnosis is evolving rapidly with encouraging findings including this study. Fahrmann et al. reported combined blood based proteomic markers with available lung cancer screening clinical tools from Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial control group (PLCOmzon) significantly improves lung cancer risk assessment while the current clinical screeningprotocols have their own strength and caveat. Our glycoproteomic platform and related biomarkers have the potential to improve upon lung cancer screening. In the pre-CT setting, optimized biomarkers could provide a non-invasive means for NSCLC screening or identifying at-risk populations. In the post-CT screening setting, with further development, glycoproteomic differential biomarkers can help to optimize nodule management and risk stratification to avoid unnecessary invasive procedures or frequent CT scans. Paramount to the future development of our glycoproteomic bio marker and classifier is our plan to pair these with clinical evaluation and risk management protocols.
[0254] In conclusion, we have presented a unique discovery study where we identified a set of glycopeptide and non-glycosylated peptide biomarkers that can differentiate patients with NSCLC from NSCLC-negative healthy and COPD controls. These biomarkers, with further development, have the potential to provide a non-invasive means for NSCLC screening that can be offered as liquid biopsies. Liquid biopsies provide an opportunity to complement current lung screening access due to their less invasive nature. Moreover, serial use of a non-invasive blood test reduces the risk of long-term radiation exposure for patients requiring follow-up.Table 12B: Peptide Structures Associated with classifying between healthy controls / COPD and NSCLC in accordance with Table 12A.Table 13: Mass Spectrometry-Related Characteristics for the Peptide Structures associated with NSCLC compared to healthy controls / COPD in accordance with Table 12BTable 14: Peptide SEQ ID NOS in accordance with Table 12BTable 15: Markers and Protein Positions in accordance with Table 12BTable 16. Sequence of Amino Acids for Proteins Corresponding to Table 12BTable 17: N-Glycan Symbol Structure GL NOS: Composition and Symbol Structure in accordance with Table 12ATable 18: O-Glycan GL NOS: Composition and Symbol Structure in accordance with Table12BLegend for Table 17 and Table 18
Claims
CLAIMS1. A method of classifying a biological sample obtained from a subject with respect to a plurality of states associated with non-small cell lung cancer (NSCLC), the method comprising receiving peptide structure data corresponding to a set of proteins in the biological sample; inputting quantification data identified from the peptide structure data for a set of peptide structures into a machine-learning model trained to identify a disease indicator based on the quantification data, wherein the set of peptide structures comprises at least one peptide structure identified from a plurality of peptide structures in Table 12B; identifying, by the machine-learning model, the disease indicator; and classifying the biological sample with respect to a plurality of states associated with NSCLC based upon the identified disease indicator.
2. A method of detecting the presence of non-small cell lung cancer (NSCLC) in a subject, the method comprising receiving peptide structure data corresponding to a set of proteins in a biological sample obtained from a subject, wherein the peptide structure data comprises at least one peptide structure from Table 12B; inputting quantification data identified from the peptide structure data for a set of peptide structures into a machine-learning model trained to identify a disease indicator based on the quantification data; and detecting the presence of NSCLC in response to a determination that the identified disease indicator falls within a selected range associated with NSCLC.
3. The method of claim 1 or 2, wherein the set of proteins comprises one or more glycoproteins.
4. The method of any one of claims 1-3, wherein the at least one peptide structure comprises a glycopeptide.
5. The method of any one of claims 1-4, wherein the plurality of states comprises at least one of a NSCLC state or a healthy state.
6. The method of any one of claims 1-5, wherein the machine-learning model comprises a regularized regression model.
7. The method of any one of claims 1-6, wherein the regularized regression model comprises a least absolute shrinkage and selection operator (LASSO) regression model.
8. The method of any one of claims 1-7, wherein the quantification data for a peptide structure of the set of peptide structures comprises at least one of an abundance, a relative abundance, a normalized abundance, or a differential abundance.
9. The method of any one of claims 1-8, wherein the quantification data for a peptide structure of the set of peptide structures comprises at least one of a relative quantity, an adjusted quantity, a normalized quantity, a relative concentration, an adjusted concentration, or a normalized concentration.
10. The method of any one of claims 1-9, wherein the quantification data is generated using a liquid chromatography-mass spectrometry (LC-MS) system.
11. The method of any one of claims 1-10, wherein the peptide structure data is generated using multiple reaction monitoring mass spectrometry (MRM-MS).
12. The method of any one of claims 1-11, wherein the machine-learning model was trained utilizing a portion of the quantification data corresponding to a set of peptide structures that is a subset of the panel of peptide structures to determine which state of the plurality of states the biological sample from the subject corresponds.
13. The method of any one of claims 1-12, further comprising: performing a differential expression analysis using the quantification data for the plurality of subjects.
14. The method of any one of claims 1-13, wherein the biological sample comprises at least one of blood, serum, or plasma.
15. A method of diagnosing an individual with non-small cell lung cancer (NSCLC), comprising: detecting the presence or amount of at least one peptide structure structures from Table 12B; inputting a quantification of the detected at least one peptide structure into a machinelearning model trained to generate a class label; determining if the class label is above or below a threshold for a classification; identifying a diagnostic classification for the individual based on whether the class label is above or below a threshold for the classification; and diagnosing the individual as having NSCLC based on the diagnostic classification.
16. The method of any one of claims 1-15, wherein the quantification data is generated using a liquid chromatography-mass spectrometry (LC-MS) system.
17. The method of any one of claims 1-16, wherein the peptide structure data is generated using multiple reaction monitoring mass spectrometry (MRM-MS).
18. The method of any one of claims 1-17, wherein the amount of at least one peptide structure is none, or below a detection limit.
19. The method of any one of claims 1-18, wherein the NSCLC is one of early-stage or late-stage NSCLC.
20. The method of any one of claims 1-19, wherein the NSCLC is one of stage I NSCLC, stage II NSCLC, stage III NSCLC, or stage IV NSCLC.
21. The method of any one of claims 1-20, wherein the at least one peptide structure comprises three or more peptide structures identified in Table 12B.
22. The method of any one of claims 1-21, wherein the at least one peptide structure comprises at least one peptide comprising the sequence set forth in any one of SEQ ID NOs: 101-142.
23. The method of any one of claims 1-22, wherein the subject has one or more risk factors associated with NSCLC.
24. The method of any one of claims 1-23, wherein the individual is determined have a healthy state, wherein a healthy state comprises the absence of NSCLC.
25. The method of any one of claims 1-24, further comprising generating a report that includes a diagnosis based on the corresponding state detected for the subject.
26. The method of any one of claims 1-25, wherein the at least one peptide structure comprises at least one, at least two, at least three, at least five, at least 10, at least 15, at least 20, or at least 25 different peptides comprising the sequence set forth in any one of SEQ ID NOs: 101-142.
27. The method of any one of claims 1-25, wherein the at least one peptide structure comprises a peptide sequence and a glycan structure, wherein the glycan structure is attached to a linking site position in the peptide sequence in accordance with Table 12B.
28. The method of claim 27, wherein the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 12B, wherein the glycan structure comprises a symbol structure in accordance with the glycan structure GL number according to Table 12B, Table 17, and Table 18.
29. The method of claim 27 or 28, wherein the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 12B, wherein the glycan structure comprises a composition in accordance with the glycan structure GL number, Table 12B, Table 17, and Table 18.
30. The method of any one of claims 1-29,wherein a rightmost N-acetylgalactosamine of the glycan structure in Table 18 is attached to a linking site position in the peptide sequence in accordance with Table 12B, and wherein a bottommost N-acetylglucosamine of the glycan structure in Table 17 is attached to a linking site position in the peptide sequence in accordance with Table 12B.
31. A composition comprising one or more peptide structures from Table 12B.
32. A composition comprising one or more peptides comprising the sequence set forth in SEQ ID NOs: 101-142.
33. The composition of claim 32, wherein the one or more peptides comprise one or more glycopeptides.
34. The composition of claim 31, wherein the at least one peptide structure comprises a peptide sequence and a glycan structure, wherein the glycan structure is attached to a linking site position in the peptide sequence in accordance with Table 12B.
35. The composition of claim 34, wherein the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 12B, wherein the glycan structure comprises a symbol structure in accordance with the glycan structure GL number according to Table 12B, Table 16, and Table 17.
36. The composition of claim 34 or 35, wherein the glycan structure of the peptide sequence corresponds to a glycan structure GL number in accordance with Table 12B, wherein the glycan structure comprises a composition in accordance with the glycan structure GL number, Table 12B, Table 16, and Table 17.
37. The composition of any one of claims 34-36, wherein a rightmost N-acetylgalactosamine of the glycan structure in Table 18 is attached to a linking site position in the peptide sequence in accordance with Table 12B, and wherein a bottommost N-acetylglucosamine of the glycan structure in Table 17 is attached to a linking site position in the peptide sequence in accordance with Table 12B.
Citation Information
Patent Citations
Genomic classification of non-small cell lung carcinoma based on patterns of gene copy number alterations
US20100145893A1
Biomarkers for diagnosing ovarian cancer
US20230055572A1
