Construction method of anti-programmed death protein-1 monoclonal antibody treatment response prediction model

Abstract

The invention provides a construction method of an anti-programmed death protein-1 monoclonal antibody treatment response prediction model, and belongs to the technical field of construction methods of prediction models. The method comprises the following steps: acquiring a blood sample of a non-small cell lung cancer patient; extracting peripheral blood mononuclear cells of the blood sample of the patient before and after the treatment of the anti-programmed death protein-1 monoclonal antibody; sequentially carrying out treatment through the steps of protein extraction, reductive alkylation and enzymolysis to obtain a sample peptide fragment; performing basic characteristic analysis and function annotation to obtain a protein spectrum and a modification spectrum of peripheral blood mononuclear cells of the non-small cell lung cancer patient treated by the anti-programmed death protein-1 monoclonal antibody; screening by utilizing machine learning to obtain a biomarker; and monitoring to obtain the anti-programmed death protein-1 monoclonal antibody treatment response prediction model. The method can be used for predicting the treatment effect of the anti-programmed death protein-1 monoclonal antibody of a non-small cell lung cancer patient.

Description

Technical Field

[0001] The present invention relates to the technical field of prediction model construction methods, and in particular to a method for constructing an anti-programmed death protein-1 monoclonal antibody treatment response prediction model. Background Art

[0002] In recent years, the discovery of immune checkpoints such as cytotoxic T lymphocyte-associated protein-4 (CTLA-4) and anti-programmed death protein-1 (PD-1) has promoted the development of tumor immunotherapy. Immune checkpoint inhibition (ICI) therapy has completely changed the treatment model of tumors by releasing the immunosuppression of T cells to stimulate a powerful anti-tumor immune response. Currently, immune checkpoint inhibitors targeting cytotoxic T lymphocyte-associated protein-4 and anti-programmed death protein-1 monoclonal antibody and its ligand (PD-L1) are increasingly used to treat various types of tumors. However, there are few studies on their therapeutic response effects. Summary of the invention

[0003] In view of this, the present invention provides a method for constructing an anti-programmed death protein-1 monoclonal antibody treatment response prediction model, which can predict the anti-programmed death protein-1 monoclonal antibody treatment response of patients with non-small cell lung cancer, thereby predicting the anti-programmed death protein-1 monoclonal antibody treatment effect of patients with non-small cell lung cancer, making it more suitable for practical use.

[0004] In order to achieve the above-mentioned object, the technical scheme of the method for constructing the anti-PD-1 monoclonal antibody treatment response prediction model provided by the present invention is as follows:

[0005] The method for constructing an anti-programmed death protein-1 monoclonal antibody treatment response prediction model provided by the present invention comprises the following steps:

[0006] Obtaining a sample from a patient with non-small cell lung cancer, wherein the patient is a sample from a patient who has been treated with an anti-programmed death protein-1 monoclonal antibody;

[0007] Extracting peripheral blood mononuclear cells from the patient sample before and after anti-PD-1 monoclonal antibody treatment, wherein for each patient sample, there is a pair of paired samples of a pre-treatment subsample and a post-treatment subsample;

[0008] The peripheral blood mononuclear cells of the patient sample before and after anti-PD-1 monoclonal antibody treatment are sequentially subjected to steps including protein extraction, reduction alkylation, and enzymatic hydrolysis to obtain sample peptides;

[0009] Processing the sample peptide segments to obtain processed sample peptide segments;

[0010] For the processed sample peptides, basic feature analysis and functional annotation are performed to obtain the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with anti-programmed death protein-1 monoclonal antibody;

[0011] Based on the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with the anti-programmed death protein-1 monoclonal antibody, machine learning is used to screen biomarkers of anti-programmed death protein-1 monoclonal antibody treatment response effects;

[0012] The biomarkers of the anti-programmed death protein-1 monoclonal antibody treatment response effect of the non-small cell lung cancer patient samples are monitored to obtain the anti-programmed death protein-1 monoclonal antibody treatment response prediction model.

[0013] The emotion recognition method provided by the present invention can also be further implemented by adopting the following technical measures.

[0014] Preferably, in the step of obtaining a sample from a patient with non-small cell lung cancer, wherein the patient is a patient treated with anti-PD-1 monoclonal antibody,

[0015] The inclusion criteria for the non-small cell lung cancer patient samples include: patients aged 18 years or older, patients clinically diagnosed with non-small cell lung cancer, patients with complete clinical information, and patients who are treated with anti-programmed death protein-1 monoclonal antibody for the first time;

[0016] The exclusion criteria for the non-small cell lung cancer patient samples include: the patient died during the treatment with anti-programmed death protein-1 monoclonal antibody, or the patient was in the treatment with anti-programmed death protein-1 monoclonal antibody.

[0017] The peripheral blood mononuclear cells of the patient sample before and after anti-PD-1 monoclonal antibody treatment are sequentially subjected to steps including protein extraction, reductive alkylation, and enzymatic hydrolysis to obtain sample peptides. The protein extraction includes the following steps:

[0018] Take out peripheral blood mononuclear cell samples from the liquid nitrogen tank, where the number of cells is 2×10 6 , thawed in a 37℃ water bath, centrifuged at 1200rpm for 3min, and discarded the cryopreservation solution.

[0019] Prepare 1× medium-acting radioimmunoprecipitation lysis buffer, wherein the volume ratio of radioimmunoprecipitation lysis buffer to protease inhibitor is 99:1,

[0020] According to the amount of peripheral blood mononuclear cells, radioimmunoprecipitation lysis buffer was added, wherein, for every 1×10 6 Add 200 μL of radioimmunoprecipitation lysis buffer to each cell.

[0021] Add two homogenate beads for lysis, homogenize once, where the frequency is 65 Hz, the total number of grinding times: 2 times, start 45 seconds and stop 10 seconds, incubate on ice for 10 minutes, repeat twice,

[0022] Centrifuge at 4°C, 12000 rpm for 10 min and remove the supernatant;

[0023] Accurately pipette 200 μg of protein solution, add 3 times the volume of ice acetone, and freeze in a -80℃ refrigerator for 1 hour.

[0024] After 1 hour, take out, centrifuge at 4℃12000rpm for 10min, discard the upper acetone, add 500μL of 80% ice acetone to the lower protein layer, vortex mix, centrifuge again under the same conditions, and discard the supernatant;

[0025] Add 200 μL of 50 mM ammonium bicarbonate solution and sonicate for 10 min on ice to redissolve the protein;

[0026] Add 2 μL of 5 mM dithiothreitol to a final concentration of 0.05 mM, vortex to mix, then centrifuge at 1000 rpm for 1 min, and place in a metal bath at 37°C for 30 min;

[0027] Add 6 μL of 5 mM iodoacetamide to a final concentration of 0.15 mM, vortex to mix, then centrifuge at 1000 rpm for 1 min, and place in a metal bath at 55°C for 30 min.

[0028] Trypsin hydrolysis: add 10 μL of 0.5 μg / μL sequencing-grade trypsin, so that the volume ratio of protein to sequencing-grade trypsin is 40:1.

[0029] Vortex to mix, then centrifuge at 1000 rpm for 1 min, and digest in a metal bath at 37°C and 900 rpm for 12 h. The next day, add 5 μL of sequencing-grade trypsin and digest again under the same conditions for 2 h.

[0030] After 2 hours, the mixture was taken out, 2 μL of 0.1% formic acid was added, vortexed for 3 seconds to terminate the reaction, and stored in a -80°C refrigerator to obtain sample peptides.

[0031] Preferably, the sample peptides are processed to obtain the processed sample peptides, which specifically comprises the following steps:

[0032] Install the octadecyl solid phase extraction column on the solid phase extraction device, and place a centrifuge tube under the column to collect the waste liquid;

[0033] Activation: Add 500 μL 100% acetonitrile to activate the octadecyl solid phase extraction column, repeat twice;

[0034] Balance: Add 500 μL 0.1% formic acid to wash away residual acetonitrile, repeat twice;

[0035] Add sample: add peptide to octadecyl solid phase extraction column, collect the effluent, repeat twice and discard the effluent;

[0036] Desalting: Wash the column with 500 μL 0.1% formic acid, repeat 2-3 times to remove as much salt as possible;

[0037] Elution of peptides: Add 200 μL of 70% acetonitrile to elute the peptides, repeat twice, and collect the eluate;

[0038] After peptide desalting, the peptides were quantified using a peptide concentration determination kit;

[0039] Turn on the low-temperature freeze dryer, set the pressure to 150kpa, and the cold trap temperature to -56℃ for precooling. After the temperature drops to the predetermined value, place the sample, open the centrifuge and vacuum valve in turn, close the atmospheric valve, and vacuum concentrate and dry the 400μL eluate collected by desalting for 3h;

[0040] After complete freeze-drying, 50 μL of 0.1% formic acid prepared in mass spectrometry water was added to dissolve the freeze-dried peptides. After sufficient vortexing, the mixture was centrifuged at 12,000 rpm at 4°C for 10 min. 45 μL of the supernatant was injected into high performance liquid preparative chromatography for fractionation to obtain the processed sample peptides.

[0041] Preferably, the HPLC fractionation method comprises the following steps:

[0042] 100× stock solution: dilute mass spectrometry grade concentrated ammonia to a concentration of 2M, take 6.7mL of 2M ammonia and add 43.3mL of mass spectrometry water, adjust the pH to 10.5, prepare 50mL of 100x stock solution and place in a 4℃ refrigerator for later use.

[0043] Aqueous buffer A: Take 485 mL of mass spectrometry water and add 10 mL of chromatography grade acetonitrile, then add 5 mL of the prepared 100× stock solution to make a mixed solution of 500 mL of 97% mass spectrometry water, 2% chromatography grade acetonitrile and 1% 100× stock solution.

[0044] Organic phase buffer B: Take 2 mL of mass spectrometry water and add 196 mL of chromatography grade acetonitrile, then add 2 mL of the prepared 100x stock solution to make a mixed solution of 200 mL of 1% mass spectrometry water, 98% chromatography grade acetonitrile and 1% 100x stock solution.

[0045] 10% methanol: Take 10 mL of chromatography grade methanol and add 90 mL of mass spectrometry water to make 10% methanol for later use.

[0046] After the required solution is prepared, tighten the bottle cap and ultrasonicate for 15 minutes to remove bubbles in the liquid;

[0047] Under the conditions of column temperature of 40℃, sample tray temperature of 7℃ and pressure of 4000psi, turn on the power supply, column oven, fraction collector, computer and UV monitor in sequence. Replace the newly prepared mobile phase, perform wet priming, flush the injector, clean the injection needle, clean the pump head and UV monitoring in sequence, and collect the absorption peaks at wavelengths of 214nm and 280nm.

[0048] Take 45 μL of sample and inject it into the chromatographic column through the automatic sample injector, then start gradient elution and collect fractions at the same time. The program of the fraction collector is set to collect one tube per minute, for a total of 56 minutes, and a total of 56 tubes are collected. The first tube is discarded, and the remaining 55 tubes are marked with serial numbers. Finally, the eluate is combined into the final 20 fractions in a certain order, concentrated by centrifugation and lyophilized to obtain the processed sample peptides.

[0049] Preferably, basic feature analysis and functional annotation are performed on the processed sample peptides to obtain the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with anti-PD-1 monoclonal antibody, which specifically includes the following steps:

[0050] The proteome discovery software (Proteome Discoverer 2.4 is used in this patent) was used to search the mass spectrometry raw data, and the 20 flow fractions of each sample were combined and matched with the human protein library downloaded from the protein database (https: / / www.uniprot.org / ) to obtain protein identification and quantification. The database search used the SEQUEST algorithm, and the parameters were set as follows: the number of amino acids constituting the peptide segment was greater than or equal to 6, a maximum of 2 missed cleavage sites were allowed, the mass tolerance of the parent ion was set at -10ppm-10ppm, and the mass tolerance of the parent ion fragment ion was -20ppm-20ppm;

[0051] The results identified by Protein Discover 2.4 were exported and the data was merged to generate a matrix. The protein quantitative value was standardized using the formula: protein abundance / (average protein of sample / average protein abundance of all samples) to eliminate the protein quantitative value errors between samples caused by sample addition amount, instrument operation, etc.

[0052] Based on the results identified by Protein Discover 2.4, after removing the contaminating proteins, the quality of mass spectrometry detection data was evaluated by peptide, protein and parent ion mass tolerance distribution analysis. The identified proteins were also subjected to basic feature analysis, annotation of functional databases including Gene Ontology and Kyoto Encyclopedia of Genes and Genomes database, and quantitative analysis, including overall differential analysis of identified proteins, screening of differentially expressed proteins, and cluster analysis of expression patterns;

[0053] Use an open search engine (OpenDelta is used in this patent) to search and match the 20 original files obtained by mass spectrometry for each sample, and download the human protein library from the protein database (https: / / www.uniprot.org / ), and use the existing modification database (https: / / www.unimod.org / ) to perform systematic annotations of known modifications, and use MSFragger software to search and annotate unknown modifications. The acetylation of the N-terminus of the protein and the oxidation on methionine are set as variable modifications, and open searches are performed for known and unknown modifications. The molecular weight deviation of the precursor ion is set at -10ppm-10ppm, and the molecular weight deviation range of the fragment ion is set at -20ppm-20ppm;

[0054] The analysis results of the open search engine (OpenDelta is used in this patent) adopt (the number of peptide matching spectra of a certain modification at a certain protein site) / (the total number of peptide matching spectra of unmodified, a certain modification, and other arbitrary modifications at the protein site), in order to remove the influence of protein abundance;

[0055] Based on the identification results of OpenDelta, in vitro modifications, including alkylation modifications introduced by pretreatment and oxidation modifications of methionine, were eliminated and potential biomarkers were screened.

[0056] Preferably, based on the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with the anti-programmed death protein-1 monoclonal antibody, in the step of screening for biomarkers of anti-programmed death protein-1 monoclonal antibody treatment response effect using machine learning, the screening conditions include:

[0057] The quantitative values ​​of each protein in two compared sample groups were subjected to T-test, and the corresponding P value was calculated as a significance indicator, and the difference fold > 2 and P value < 0.05 were used as the conditions for screening differentially expressed proteins;

[0058] The quantitative values ​​of each modification in two compared sample groups were subjected to a rank sum test, and the corresponding P value was calculated as a significance indicator. The P value was corrected using the false discovery rate method, and the difference fold > 2 and P value < 0.05 were used as the conditions for screening differentially expressed proteins.

[0059] The random forest method was used for simulation screening to further explore candidate biomarkers that can be used to predict the benefits of anti-programmed death protein-1 monoclonal antibody treatment. In the random forest analysis, 50,000 trees were constructed using the R package random forest, and the maximum number of terminal node trees in the forest was set to 5. In order to evaluate the effectiveness of the model, the subject survival characteristic curve was drawn, and the area under the curve (AUC) of its 95% confidence interval was calculated.

[0060] Preferably, based on the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with the anti-programmed death protein-1 monoclonal antibody, in the step of screening for biomarkers of anti-programmed death protein-1 monoclonal antibody treatment response effects using machine learning, the analysis method comprises:

[0061] Cluster analysis of protein expression patterns: cluster analysis of differentially expressed proteins was performed using heat map and fuzzy mean clustering based pseudo-time sequence analysis (MFUZZ) methods;

[0062] Gene Ontology and Kyoto Encyclopedia of Genes and Genomes functional enrichment analysis: The ClueGO plug-in within the Cytoscape3.8.1 software was used to perform functional and pathway enrichment analysis on differentially expressed proteins;

[0063] Protein interaction network analysis: The target proteins were imported into the STRING database to obtain protein interaction relationships and the results were imported into Cytoscape 3.8.1 software to draw the protein interaction network diagram. The Cytoscape internal plug-ins Mcode and cytoHubba were used to obtain the corresponding core modules and core nodes respectively.

[0064] Preferably, based on the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with the anti-programmed death protein-1 monoclonal antibody, in the step of screening for biomarkers of anti-programmed death protein-1 monoclonal antibody treatment response effects using machine learning, the verification method comprises:

[0065] The protein pretreatment of the validation cohort included protein extraction, reduction, alkylation, enzymatic hydrolysis, and desalting;

[0066] Proteome: Select peptides of target proteins according to the following criteria to establish parallel reaction monitoring methods: 1) Only select peptides unique to proteins; 2) No miscleavage sites in peptides; 3) No variable modification sites in peptides, such as oxidation and alkylation; 4) Peptides that meet the above requirements are selected with high response values; 5) At least one and up to three peptides unique to proteins are selected for each protein;

[0067] Modification group: Parallel reaction monitoring method is established for target modified peptides: 1) There are no mis-cleavage sites in the peptides; 2) There are no variable modification sites in the peptides, such as oxidation modification and alkylation modification; 3) Peptides that meet the above requirements are selected with high response values;

[0068] The samples were analyzed using a reversed-phase octadecyl self-packed capillary column (75 μm×100 mm) and separated using an EASY-nLC 1200 ultra-high performance liquid chromatography system, wherein mobile phase A was an aqueous solution containing 0.1% formic acid and 2% acetonitrile; mobile phase B was an aqueous solution containing 0.1% formic acid and 90% acetonitrile; the gradient was set as follows: 0-46 min, 6% to 28% B phase; 46-53 min, 28%-45% B phase; 53-60 min, 100% B phase, the flow rate was maintained at 350 nL / min, and the column temperature was 50°C). The peptides were separated by ultra-high performance liquid phase system and injected into a sodium spray ion source (voltage 2.1 kV) and detected by a ThermoScientificTM Q ExactiveHF-X (Thermo Fisher) mass spectrometer, wherein the mass spectrometry data were acquired in a high-sensitivity parallel reaction monitoring mode with the following parameters: high-energy collision dissociation, collision energy 32%, primary resolution 60,000, secondary resolution 15,000, scanning range 100-1800 m / z, scanning time 100 ms, screening for charge states of +2-+4 including precursors, dynamic exclusion time 30 s, isolation window 4 Da, time window 7 min;

[0069] Skyline 3.6 software was used for parallel reaction monitoring data processing. All results were imported into Skyline, and the correct peaks were selected to obtain the results of all peptides in all samples. Among them, the parallel reaction monitoring results, including protein names and peak areas, were further exported for analysis to screen and compare differential proteins in different groups.

[0070] Based on the relative quantitative results obtained by the Skyline peak area, the R language was used for correlation analysis and visualization of the differences, and the R package was used to draw the receiver operating characteristic (ROC) curve and calculate the area under the curve.

[0071] Preferably, in the step of monitoring the biomarkers of the anti-programmed death protein-1 monoclonal antibody treatment response effect of the non-small cell lung cancer patient samples and obtaining the anti-programmed death protein-1 monoclonal antibody treatment response prediction model, the non-small cell lung cancer patient samples are classified into three subtypes: prediction group, resistance group, and benefit group according to the expression trends of modular proteins and modifications.

[0072] The method for constructing a model for predicting the therapeutic response to anti-programmed death protein-1 monoclonal antibody provided by the present invention uses a pre-treatment subsample and a post-treatment subsample as a pair of paired samples, then performs protein extraction, reduction alkylation, and enzymatic hydrolysis steps on peripheral blood mononuclear cells of the patient samples before and after the anti-programmed death protein-1 monoclonal antibody treatment to obtain sample peptide segments, and after further processing, obtains a biomarker of the therapeutic response effect of the anti-programmed death protein-1 monoclonal antibody, and then monitors the biomarker to obtain a model for predicting the therapeutic response to anti-programmed death protein-1 monoclonal antibody, which can predict the therapeutic effect of the anti-programmed death protein-1 monoclonal antibody for patients with non-small cell lung cancer. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0074] Figure 1 A flowchart of the steps of the method for constructing a prediction model for anti-programmed death protein-1 monoclonal antibody treatment response provided in an embodiment of the present invention;

[0075] Figure 2 This is a graph showing the protein molecular subtype characteristics (K-means clustering algorithm) of patients with non-small cell lung cancer treated with anti-programmed death protein-1 monoclonal antibody;

[0076] Figure 3 A is a variable importance map generated by the random forest algorithm for the screening of potential response markers for anti-programmed death protein-1 monoclonal antibody treatment, measuring the average reduced Gini coefficient of the accuracy of each variable between responding and non-responding patients before treatment;

[0077] Figure 3 B is the area under the curve (top 20) for screening potential response markers for anti-programmed death protein-1 monoclonal antibody treatment, which is used to evaluate the ability of a single protein to distinguish between responding and non-responding patients before treatment;

[0078] Figure 3 C is a Venn diagram for screening potential response markers for anti-programmed death protein-1 monoclonal antibody treatment, showing the intersection of 154 protein maps;

[0079] Figure 3 D is a graph of 20 key characteristic proteins (yellow indicates acceptable) obtained by random forest combined with Boruta feature screening for potential response markers for anti-programmed death protein-1 monoclonal antibody treatment;

[0080] Figure 4The alternative protein interaction network can be used to predict whether anti-programmed death protein-1 monoclonal antibody treatment is effective; the colors in the figure represent the protein difference fold (before receiving anti-programmed death protein-1 monoclonal antibody treatment, the protein expression level in peripheral blood mononuclear cells of the responding group was lower than that of the non-responding group); the thickness of the line indicates whether the protein expression levels are correlated (Spearman correlation analysis), the thicker the line, the more correlated; the size of the circle represents the area under the curve value of each protein;

[0081] Figure 5 A is the receiver operating characteristic curve of a single indicator of the five protein clusters as single indicators and the receiver operating characteristic curve of the combined indicators;

[0082] Figure 5 B is the receiver operating characteristic curve of the five protein clusters as single indicators and combined indicators predicted by the receiver operating characteristic curve of multiple indicators stepwise combined;

[0083] Figure 6 Receiver operating characteristic curves for the remaining 16 as single indicators;

[0084] Figure 7 A is the ubiquitin modification profile of peripheral blood mononuclear cells of non-small cell lung cancer patients treated with anti-programmed death protein-1 monoclonal antibody (distribution of the number of peptide matching patterns of known modifications (chemical and biological modifications) and unknown modifications);

[0085] Figure 7 B is the ubiquitin modification profile of peripheral blood mononuclear cells of non-small cell lung cancer patients treated with anti-programmed death protein-1 monoclonal antibody (the number of peptide matching spectra of each subclass of biological modification);

[0086] Figure 7 C is the ubiquitin modification profile of peripheral blood mononuclear cells of non-small cell lung cancer patients treated with anti-programmed death protein-1 monoclonal antibody (the profile of the number of peptide matching spectra of each subcategory of oxidative stress modification);

[0087] Figure 7 D is the ubiquitin modification profile of peripheral blood mononuclear cells of non-small cell lung cancer patients treated with anti-programmed death protein-1 monoclonal antibody (the profile of the number of peptide matching spectra of each subcategory of metabolite modification);

[0088] Figure 8 This is a diagram of the modified molecular subtype characteristics (K-means clustering algorithm) of patients with non-small cell lung cancer after treatment with anti-programmed death protein-1 monoclonal antibody;

[0089] Fig. 9 The receiver operating characteristic curve is used as a single indicator prediction;

[0090] Fig.10A is the receiver operating characteristic curve (the receiver operating characteristic curve of a single indicator) of the five protein clusters screened in the discovery set as single indicators and combined indicators prediction using parallel reaction monitoring test results in the validation cohort;

[0091] Fig.10 B is the receiver operating characteristic curve (the receiver operating characteristic curve of multiple indicators stepwise combined) of the five protein clusters screened in the discovery set as single indicators and combined indicators for prediction, using parallel reaction monitoring test results in the validation cohort. DETAILED DESCRIPTION

[0092] In view of this, the present invention provides a method for constructing an anti-programmed death protein-1 monoclonal antibody treatment response prediction model, which can predict the anti-programmed death protein-1 monoclonal antibody treatment response of patients with non-small cell lung cancer, thereby predicting the anti-programmed death protein-1 monoclonal antibody treatment effect of patients with non-small cell lung cancer, making it more suitable for practical use.

[0093] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the construction method of a prediction model for anti-programmed death protein-1 monoclonal antibody treatment response proposed by the present invention, its specific implementation method, structure, characteristics and effects as follows in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form.

[0094] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships may exist, for example, A and / or B. Specifically, it is understood that: A and B may be included at the same time, A may exist alone, or B may exist alone, and any of the above three situations may be met.

[0095] Peripheral blood mononuclear cell samples of patients with non-small cell lung cancer were collected before and after two cycles (6 weeks) of treatment with anti-PD-1 monoclonal antibodies (sintilimab, camrelizumab, and toripalimab) (Table 1). According to clinical criteria, patients with partial remission (PR, 9 cases) and stable disease (SD, 28 cases) on the first computed tomography scan after 6 weeks of treatment were classified as responders (Rs) to anti-PD-1 monoclonal antibody treatment; patients with progressive tumors on the first computed tomography scan and patients with progressive disease (PD, 16 pairs) within 6 weeks after the start of treatment were classified as non-responders (NRs) to anti-PD-1 monoclonal antibody treatment.

[0096] The inclusion and exclusion criteria for the samples were as follows:

[0097] (1) Aged 18 or above;

[0098] (2) pathological diagnosis of non-small cell lung cancer;

[0099] (3) Patients who were treated with anti-programmed death protein-1 monoclonal antibody for the first time were enrolled in the group, and those who died or changed drugs during treatment were excluded from the group;

[0100] (4) Have complete clinical information.

[0101] Table 1 Clinical baseline data of patients in the discovery cohort

[0102]

[0103] 1 :n(%); median.

[0104] 2 : Pearson chi-square test; Kruskal-Wallis test; Fisher's exact probability test.

[0105] Protein extraction and enzymatic hydrolysis

[0106] 1. Take out the peripheral blood mononuclear cell sample from the liquid nitrogen tank (the number of cells is about 2 x 10 6 ), thawed in a 37°C water bath, centrifuged at 1200 rpm for 3 min, and the cryopreservation solution was discarded;

[0107] 2. Prepare 1x medium-acting radioimmunoprecipitation lysis buffer (radioimmunoprecipitation lysis buffer: protease inhibitor = 99:1);

[0108] 3. Add radioimmunoprecipitation lysis buffer (1x 10 6 Add approximately 200 μL of radioimmunoprecipitation lysis buffer per cell).

[0109] 4. Add two homogenized beads for lysis, homogenize once (65 Hz, total grinding times: 2 times, 45 s on and 10 s off), incubate on ice for 10 min, repeat twice;

[0110] 5. Centrifuge at 4°C, 12000 rpm for 10 min and remove the supernatant;

[0111] 6. The protein concentration was determined by the bicinchoninic acid (BCA) method.

[0112] 7. Accurately pipette 200 μg of protein solution, add 3 times the volume of ice acetone, and freeze in a -80℃ refrigerator for 1 hour;

[0113] 8. Take out after 1 hour, centrifuge at 4℃12000rpm for 10min, discard the upper acetone layer, add 500μL 80% ice acetone to the lower protein layer, vortex mix, centrifuge again under the same conditions, and discard the supernatant;

[0114] 9. Add 200 μL of 50 mM ammonium bicarbonate solution and sonicate for 10 min (ice bath) to redissolve the protein;

[0115] 10. Add 2 μL of 5 mM dithiothreitol (final concentration 0.05 mM), vortex to mix, then centrifuge at low speed for a short time (1000 rpm, 1 min), and place in a metal bath at 37°C for 30 min;

[0116] 11. Add 6 μL of 5 mM iodoacetamide (final concentration 0.15 mM), vortex to mix, then centrifuge at low speed for a short time (1000 rpm, 1 min), and place in a metal bath at 55°C for 30 min;

[0117] 12. Trypsin hydrolysis: add 10 μL sequencing grade trypsin (0.5 μg / μL, protein: trypsin = 40:1),

[0118] Vortex to mix, then centrifuge briefly at low speed (1000 rpm, 1 min), and digest in a constant temperature shaking metal bath (37°C 900 rpm) for 12 h. Add 5 μL of trypsin the next day and digest again under the same conditions for 2 h.

[0119] 13. Take out after 2 hours, add 2 μL 0.1% formic acid, vortex for 3 seconds to terminate the reaction, and freeze in a -80℃ refrigerator.

[0120] Peptide desalting

[0121] 1. Install the octadecyl solid phase extraction column on the solid phase extraction device and place a centrifuge tube under the column to collect the waste liquid.

[0122] 2. Activation: Add 500 μL 100% acetonitrile to activate the desalting column and repeat twice.

[0123] 3. Balance: Add 500 μL 0.1% formic acid to wash away residual acetonitrile and repeat twice.

[0124] 4. Add sample: Add the peptide to the octadecyl solid phase extraction column, collect the effluent, repeat twice and discard the effluent.

[0125] 5. Desalting: Wash the column with 500 μL 0.1% formic acid, repeat 2 to 3 times to remove as much salt as possible.

[0126] 6. Elution of peptides: Add 200 μL 70% acetonitrile to elute the peptides, repeat twice, and collect the eluate.

[0127] Peptide Quantification

[0128] After peptide desalting, the peptide concentration was quantified using a peptide concentration determination kit.

[0129] Vacuum concentration and centrifugal drying

[0130] Turn on the low-temperature freeze dryer, set the pressure to 150 kPa, and the cold trap temperature to -56 °C for precooling. After the temperature drops to the predetermined value, place the sample, open the centrifuge and vacuum valve in turn, close the atmospheric valve, and vacuum concentrate and dry the 400 μL eluate collected by desalting for 3 hours.

[0131] Peptide reconstitution

[0132] After complete freeze-drying, 50 μL of 0.1% formic acid prepared in mass spectrometry water was added to dissolve the freeze-dried peptides. After thorough vortexing, the mixture was centrifuged at 12,000 rpm for 10 min at 4°C. 45 μL of the supernatant (approximately 200 μg of peptide) was taken and injected into high performance liquid chromatography (HPLC) for fractionation.

[0133] HPLC fractionation

[0134] Since peripheral blood mononuclear cells contain many types of cells, the extracted protein components are also more complex. Therefore, the obtained peptide samples are fractionated by HPLC preparative chromatography and then tested by LC-MS to obtain deeper protein and modification spectrum information.

[0135] HPLC preparative chromatography fractionation is to load the peptides onto a specific chromatographic column via a mobile phase, and use different proportions of buffer A and buffer B to perform high alkalinity reverse phase elution to achieve the purpose of separating the peptides. The elution gradient is as follows:

[0136]

[0137] The specific steps of HPLC preparative chromatography fractionation are as follows:

[0138] Prepare the liquid phase:

[0139] 100x stock solution: dilute mass spectrometry grade concentrated ammonia water to a concentration of 2M, take 6.7mL of 2M ammonia water and add 43.3mL of mass spectrometry water, adjust the pH to 10.5, prepare 50mL of 100x stock solution and store in a 4℃ refrigerator for later use.

[0140] Buffer A (aqueous phase): Take 485 mL of mass spectrometry water and add 10 mL of chromatography grade acetonitrile (acetonitrile), then add 5 mL of the prepared 100x stock solution to make a mixed solution of 500 mL of 97% mass spectrometry water, 2% chromatography grade acetonitrile (acetonitrile) and 1% 100x stock solution.

[0141] Buffer B (organic phase): Take 2 mL of mass spectrometry water and add 196 mL of chromatography grade acetonitrile (acetonitrile), then add 2 mL of the prepared 100x stock solution to make a mixed solution of 200 mL of 1% mass spectrometry water, 98% chromatography grade acetonitrile (acetonitrile) and 1% 100x stock solution.

[0142] 10% methanol: Take 10 mL of chromatography grade methanol and add 90 mL of mass spectrometry water to make 10% methanol for later use.

[0143] After the required solution is prepared, tighten the bottle cap and ultrasonicate for 15 minutes to remove bubbles in the liquid.

[0144] Setting parameters:

[0145] Column temperature 40°C, sample plate temperature 7°C, pressure 4000 psi.

[0146] Grading Operations:

[0147] Turn on the power supply, column oven, fraction collector, computer and UV monitor in sequence. Replace the newly prepared mobile phase, perform wet priming, flush the injector, clean the injection needle, clean the pump head and UV monitoring in sequence, and collect the absorption peaks at 214nm and 280nm.

[0148] Take 45 μL of sample and inject it into the chromatographic column through the automatic sample injector, then start gradient elution and collect fractions at the same time. The program of the fraction collector is set to collect one tube per minute, for a total of 56 minutes, and a total of 56 tubes are collected. The first tube is discarded, and the remaining 55 tubes are marked with serial numbers. Finally, the eluate is combined into the final 20 fractions in a certain order, and then centrifuged, concentrated and freeze-dried.

[0149] LC-MS detection

[0150] In this experiment, the Orbitrap Fusion Lumos (Thermo Fisher Scientific) ultra-high resolution protein spectrometer was used in combination with the Easy-nLC1200 nanoliter liquid chromatography system to collect signals of the fractionated peptides.

[0151] The chromatographic column uses octadecyl as filler, with a column length of 15 cm, an inner diameter of 75 μm, and a filler particle size of 1.9 μm. The mobile phase is prepared as follows: Phase A: 0.1% formic acid; Phase B: 80% acetonitrile, 0.1% formic acid. The time for each sample to be eluted through the liquid phase is 60 minutes, and the elution gradient is as follows:

[0152]

[0153] High-resolution mass spectrometry detection uses high-energy collision dissociation (HCD) fragmentation with a fragmentation energy of 28%. The mass analyzer is an orbital ion trap, and data-dependent acquisition is performed (DDA). The primary scan (MS1) resolution is set to 60,000, and the automatic gain control target is 3e 6 The maximum ion injection time was 50 ms, and the scanning range was 100-1800 mass-to-charge ratio (m / z); the secondary scan (MS2) selected the top 20 parent ions in MS1 ​​for further fragmentation, with a resolution of 15000 and an automatic gain control target of 5e 4 , the maximum ion injection time is 45ms and the dynamic exclusion time is 50s.

[0154] Candidate biomarker screening

[0155] Proteomic data analysis

[0156] Protein Discover 2.4 software was used to search the mass spectrometry raw data. The 20 fractions of each sample were combined and matched with the human protein library downloaded from the protein database (https: / / www.uniprot.org / ). The parameters were set as follows: the number of amino acids constituting the peptide segment was greater than or equal to 6, a maximum of 2 missed cleavage sites were allowed, the mass tolerance of the parent ion was set to -10ppm-10ppm, and the mass tolerance of the fragment ion was set to -20ppm-20ppm.

[0157] Protein data normalization

[0158] The results identified by Protein Discover 2.4 were exported and the data was merged to generate a matrix. The protein quantitative value was standardized using quantile, that is, protein abundance / (average abundance of samples / average abundance of all samples) to eliminate the protein quantitative value errors between samples caused by sample addition, instrument operation, etc.

[0159] Proteomic data screening

[0160] Based on the results identified by Protein Discover 2.4, after removing the contaminating proteins, the quality of mass spectrometry detection data was evaluated by peptide, protein and parent ion mass tolerance distribution analysis. The identified proteins were also subjected to basic feature analysis, annotation of functional databases including Gene Ontology and Kyoto Encyclopedia of Genes and Genomes database, and quantitative analysis, including overall differential analysis of identified proteins, screening of differentially expressed proteins, and cluster analysis of expression patterns.

[0161] Modified group data analysis

[0162] OpenDelta was used to search and match the spectra of the 20 raw files obtained by mass spectrometry for each sample with the human protein library downloaded from the protein database (https: / / www.uniprot.org / ), and the existing modification database (https: / / www.unimod.org / ) was used to systematically annotate known modifications, and the MSFragger software was used to search and annotate unknown modifications. The acetylation of the N-terminus of the protein and the oxidation on methionine were set as variable modifications, and an open search was performed for known and unknown modifications. The molecular weight deviation range of the precursor ion was set at -10ppm-10ppm, and the molecular weight deviation range of the fragment ion was set at -20ppm-20ppm;

[0163] Data pre-processing and standardization

[0164] The analysis results of the open search engine (OpenDelta is used in this patent) adopt (the number of peptide matching spectra of a certain modification at a certain protein site) / (the total number of peptide matching spectra of unmodified, a certain modification and other arbitrary modifications at the protein site) in order to remove the influence of protein abundance.

[0165] Modification group data screening

[0166] Based on the identification results of OpenDelta, in vitro modifications, including alkylation modifications introduced by pretreatment and oxidation modifications of methionine, were eliminated and potential biomarkers were screened.

[0167] Biomarker Screening

[0168] Statistical analysis

[0169] Differential expression analysis uses the ratio of the mean of all biological measurement values ​​of each protein (modification) in two compared sample groups as the fold change (FC). In order to determine the statistical significance of the difference, the quantitative value of each protein in the two compared sample groups was subjected to a T test, and the corresponding P value was calculated as a significance indicator, and the fold change > 2 and P value < 0.05 were used as the conditions for screening differential proteins.

[0170] The quantitative values ​​of each modification in two compared sample groups were subjected to a rank sum test, and the corresponding P value was calculated as a significance indicator. The P value was corrected using the false discovery rate method, and the difference fold > 2 and P value < 0.05 were used as the conditions for screening differentially expressed proteins.

[0171] Random forest method was used for simulation screening to further explore candidate biomarkers that can be used to predict the benefit of anti-programmed death protein-1 monoclonal antibody treatment. In the random forest analysis, 50,000 trees were constructed using the R package random forest (version 4.5.14), and the maximum number of terminal node trees in the forest was set to 5. To evaluate the effectiveness of the model, the receiver operator characteristic (ROC) curve was plotted, and the area under the curve (AUC) of its 95% confidence interval was calculated.

[0172] Cluster analysis of protein expression patterns

[0173] The differentially expressed proteins were clustered using heatmap and fuzzy mean clustering based pseudo-time sequence analysis (MFUZZ) methods.

[0174] Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis

[0175] The ClueGO plug-in within Cytoscape 3.8.1 software was used to perform gene ontology and Kyoto Encyclopedia of Genes and Genomes functional analysis and enrichment analysis on the differentially expressed proteins, mainly including biological process and pathway enrichment analysis.

[0176] Protein interaction network analysis

[0177] The target proteins were imported into the STRING website (https: / / cn.string-db.org / ) to obtain protein interaction relationships and the results were imported into Cytoscape 3.8.1 software to draw the protein interaction network diagram. The Cytoscape internal plug-ins Mcode and cytoHubba were used to obtain the corresponding core modules and core nodes, respectively.

[0178] Verification Queue

[0179] Study subjects

[0180] This study used peripheral blood mononuclear cell samples of 100 patients with non-small cell lung cancer who received anti-programmed death protein-1 monoclonal antibody (sintilimab, camrelizumab, toripalimab) before treatment (Table 2). This cohort only selected peripheral blood mononuclear cell samples of non-small cell lung cancer patients before receiving anti-programmed death protein-1 monoclonal antibody treatment.

[0181] Table 2 Clinical baseline data of patients in the validation cohort

[0182]

[0183] 1 :n(%); median.

[0184] 2 : Pearson chi-square test; Kruskal-Wallis test; Fisher's exact probability test.

[0185] Parallel reaction monitoring sample preparation

[0186] The protein pretreatment of the validation cohort was the same as that in section 1.1, except that HPLC preparative chromatography fractionation was not performed. The rest of the protein extraction, reduction, alkylation, enzymatic hydrolysis, and desalting operations were the same as those in section 1.1.

[0187] Establishment of parallel reaction monitoring method

[0188] Proteome: The peptides of the target protein were selected according to the following criteria to establish a parallel reaction monitoring method: 1) Only peptides unique to the protein were selected; 2) There were no miscleavage sites in the peptides; 3) There were no variable modification sites in the peptides, such as oxidation modification and alkylation modification; 4) Peptides that met the above requirements were selected with high response values; 5) At least one and up to three peptides unique to each protein were selected.

[0189] Modification group: Establish parallel reaction monitoring method for target modified peptides: 1) There are no mis-cleavage sites in the peptides; 2) There are no variable modification sites in the peptides, such as oxidation modification and alkylation modification; 3) Peptides that meet the above requirements are selected with high response values.

[0190] Parallel Reaction Monitoring Method

[0191] To ensure the quality of the data, the analysis of the mixed samples was used as a quality control sample to observe the stability of the instrument signal during the entire analysis process. To avoid systematic errors, samples from different groups were subjected to mass spectrometry analysis in a random order. The samples were analyzed using a reversed-phase octadecyl self-packed capillary column (75 μm × 100 mm) and separated using an EASY-nLC 1200 ultra-high performance liquid phase system (mobile phase A was an aqueous solution containing 0.1% formic acid and 2% acetonitrile; mobile phase B was an aqueous solution containing 0.1% formic acid and 90% acetonitrile; the gradient was set as follows: 0-46 min, 6% to 28% B phase; 46-53 min, 28% to 45% B phase; 53-60 min, 100% B phase, the flow rate was maintained at 350 nL / min, and the column temperature was 50 °C). The peptides were separated by ultra-high performance liquid phase system and injected into the sodium spray ion source (voltage 2.1 kV) and detected by ThermoScientificTM Q Exactive HF-X (Thermo Fisher Scientific Co., Ltd.) mass spectrometer. The mass spectrometry data were acquired in high-sensitivity parallel reaction monitoring mode with the following parameters: high-energy collision dissociation, collision energy of 32%, primary resolution of 60,000, secondary resolution of 15,000, scanning range of 100-1800 m / z, scanning time of 100 ms, screening of charge states of +2 to +4 including precursors, dynamic exclusion time of 30 s, isolation window of 4 Da, and time window of 7 min.

[0192] Parallel reaction monitoring data analysis

[0193] Skyline 3.6 software was used for parallel reaction monitoring data processing. All results were imported into the software, and the correct peaks were manually selected to obtain the results of all peptides in all samples. The parallel reaction monitoring results, including protein names and peak areas, were further exported for analysis to screen and compare differential proteins in different groups. Existing studies did not use internal reference proteins, and the normalization results were not good, so the unnormalized peptide abundance was used; if there is time, you can try to normalize the abundance of each peptide in each sample, and use the total ion chromatogram of each sample to correct the sample loading and mass spectrometry signal intensity.

[0194] Based on the relative quantitative results obtained by Skyline peak area, R language was used for correlation analysis and visualization of the differences, and R package was used to draw the receiver operating characteristic curve and calculate the area under the curve (AUC).

[0195] result

[0196] Screening of candidate biomarkers for anti-programmed death protein-1 monoclonal antibody in the treatment of non-small cell lung cancer based on proteomics

[0197] In this study, a total of 11,531 high-confidence proteins were identified using PD software, including 77 contaminating proteins. After removing the contaminating proteins, 11,454 proteins remained. These proteins were analyzed and screened to obtain potential biomarkers that can be used to predict whether anti-programmed death protein-1 monoclonal antibody treatment is effective. Proteomics-based anti-programmed death protein-1 monoclonal antibody monoclonal antibody treatment of non-small cell lung cancer molecular subtypes

[0198] After summarizing all differentially expressed proteins (a total of 1152), the K-means clustering algorithm was used to obtain six characteristic modules: Figure 1 ). Combined with clinical data, these module proteins and modifications can be classified into three subtypes according to their expression trends: “prediction group”, “resistance group” and “benefit group”.

[0199] Before receiving anti-PD-1 monoclonal antibody treatment, the protein expression level of module A in the peripheral blood mononuclear cells of the responder group was significantly higher than that in the non-responder group, while PD1-monoclonal antibody treatment had no significant regulatory effect on the protein expression level of module A; the protein expression trend of module D was similar to that of module A, and PD1-monoclonal antibody treatment had no significant regulatory effect on the protein expression level of this module; however, before receiving anti-PD-1 monoclonal antibody treatment, the expression levels of the proteins of these two modules in the peripheral blood mononuclear cells of the responder group were significantly lower than those in the non-responder group. In summary, the proteins of modules A and D can be used to predict whether patients can benefit from anti-PD-1 monoclonal antibody treatment, so they are classified as the "prediction group" subtype.

[0200] Screening of potential biomarkers for predicting response based on random forest combined with Boruta signature

[0201] In order to screen potential biomarkers related to response, the proteins in the "prediction group" and the proteins with significant differences (p<0.05, abs(log difference fold)>=1) between the response group and the non-response group before treatment were first screened out; then the gini value of each protein was obtained by combining machine learning (random forest) Figure 2 A); draw the receiver operating characteristic curve of each protein and obtain the corresponding area under the curve ( Figure 2 B); Combining these three indicators, 154 proteins were initially screened to predict whether patients can benefit from anti-programmed death protein-1 monoclonal antibody treatment ( Figure 2 C). Random forest combined with Boruta features was further used to identify 20 key characteristic proteins in the pre-treatment response group and non-response group ( Figure 2 D).

[0202] In summary, potential biomarkers for predicting whether patients can benefit from anti-PD-1 monoclonal antibody treatment can be obtained. By constructing the interaction network of these response proteins, five highly connected core protein clusters can be obtained ( Figure 3 ) and 10 core proteins. The core proteins were mainly distributed in core protein cluster 1 (Cluster 1). The first protein cluster was enriched with CDC42, ROCK1, RHOG, ARF6, RHOA, CYBA, CYBB, MMP9, S100A8 and S100A9; the second protein cluster was enriched with LSM7, LSM3, LRMDA and SNRPD3; the third protein cluster was enriched with PKM, PGAM1, HK2 and FBP1 related to sugar metabolism; the fourth protein cluster included RETN, MMP8 and DEFA1B; the fifth protein cluster included GM2A, GBA and ASAH1.

[0203] Anti-PD-1 monoclonal antibody therapy may exert its therapeutic effect by regulating the energy metabolism process of immune cells. The above screening results also show that potential biomarkers that can be used to predict whether patients can benefit from anti-PD-1 monoclonal antibody therapy are also mainly concentrated on mitochondrial energy metabolism and glucose metabolism-related proteins. Among them, the proteins that affect the functional state of immune cells are mainly distributed in protein cluster 1. RHOG downregulation promotes T cell receptor signaling and proliferation, and can affect late T cell activation; RHOA plays an important role in the formation of immune synapses between dendritic cells and antigen-specific T cells; in tumor cells, abnormal activation of ARF6 is a potential therapeutic target. When the cell energy is sufficient, ARF6 can regulate the distribution of mitochondria in the cell and protect mitochondria from oxidative damage; neutrophils, as the most abundant innate immune cells in the circulation, are the first immune cells to be recruited. S100A8 / A9 can induce neutrophil activation and enhance immune response. PGAM1 in protein cluster 3 is a glucose metabolism-related protein. Inhibition of PGAM1 can promote CD8+T cell infiltration and synergize with anti-PD-1 immunotherapy to exert anti-tumor effects. RETN in protein cluster 4 can improve the chemotaxis of immune cells and facilitate tumor immune escape; DEFA1B belongs to the α-defensin subgroup and can inhibit the expression of cytokines through CD8 + The chemotactic properties of T cells affect the tumor microenvironment, triggering the degranulation of neutrophils and thus exerting a strong anti-tumor immune response. Therefore, these proteins can affect the energy metabolism of mitochondria and the functional state of immune cells, thereby affecting the body's response to anti-programmed death protein-1 monoclonal antibody treatment. This shows that these proteins have the potential to be used to predict whether patients with non-small cell lung cancer can benefit from anti-programmed death protein-1 monoclonal antibody treatment, but their predictive efficacy needs to be further studied in new clinical cohorts.

[0204] Proteomic Biomarker Validation

[0205] We further compared the receiver operating characteristic curves of individual proteins in these protein clusters and the joint receiver operating characteristic curves ( Figure 4 ), a combined marker model that can be used to predict whether anti-programmed death protein-1 monoclonal antibody treatment is effective. In the first protein cluster, the highest area under the curve value is RHOG (0.82), followed by CYBB (0.81), and the lowest is CYBA, with an area under the curve value of only 0.65. When RHOG and CYBB are analyzed together, the area under the curve value is increased to 0.83. After gradually accumulating the protein, the area under the curve value gradually increases. Finally, using RHOG, CYBB, CDC42, RHOA, ROCK1, ARF6, S100A8 and S100A9 as joint indicators, a prediction model with an area under the curve value of 1 can be obtained. The area under the curve values ​​of LSM3, LSM7, LRMDA and SNRPD3 in protein cluster 2 range from 0.66 to 0.82, and after the protein is gradually accumulated, the area under the curve value gradually increases from 0.82 to 0.88. Protein clusters 3, 4, and 5 also show the same trend. It can be preliminarily concluded that these proteins can be used as combined diagnostic markers with higher diagnostic sensitivity.

[0206] Therefore, we preliminarily obtained a combination of markers that can be used to predict whether anti-PD-1 monoclonal antibody treatment is effective, namely, cluster 1 (RHOG, CYBB, CDC42, RHOA, ROCK1, ARF6, S100A8 and S100A9), cluster 2 (LSM3, SNRPD3), cluster 3 (FBP1, PGAM1, PKM), cluster 4 (DEFA1, MMP8, RETN) and cluster 5 (ASAH1, GM2A) as combined indicators; ARF6, EEF1D, EIF2S3, FARSB, GCA, GNG2, H4C9, LAMP1, LMAN1, MNDA, NAMPT, NDUFS3, PSMC4, PSMD6, RAB21, RAB35 and RNASE2 as single indicators of anti-PD-1 monoclonal antibody treatment response prediction model ( Figure 5 ).

[0207] Screening of candidate biomarkers for anti-programmed death protein-1 monoclonal antibody treatment of non-small cell lung cancer based on ubiquitin modification genomics

[0208] This study used an open search engine to map the ubiquitin modification map in peripheral blood mononuclear cells of patients with non-small cell lung cancer who were treated with anti-programmed death protein-1 monoclonal antibodies. A total of 804 modifications were identified in the peripheral blood mononuclear cell modification data of patients with non-small cell lung cancer, occurring at 92,937 sites on 9,013 proteins, and the total number of peptide matching maps (PSMs) for these modifications was 2,952,321.

[0209] According to the mass shift of each modification, the amino acid (AA) and the position of the amino acid in the peptide, the existing modification database (https: / / www.unimod.org / ) was used to systematically annotate it. According to the annotation results, 721 modifications from 2 923 027 (99%) peptide matching spectra were known modifications; the remaining 29294 peptide matching spectra contained 85 peptides with unknown mass shifts, which may be peptides containing new modifications, accounting for 1.0% of the total number of identified peptide matching spectra.

[0210] According to the sources of known modifications, 721 known modifications derived from 2,923,027 (99%) peptide matching maps were further divided into chemical modifications (2,128,220 PSMs, 72.00%, Figure 6 A) and biological modifications (512, 794807PSMs, 27%). A total of 211 known chemical modifications were identified in this data, including derivative modifications caused by sample pretreatment, such as propionaldehyde (+40.0313Da, 540908PSMs) caused by acetone precipitation, aminomethylation (+57.0215Da, 9353PSMs) produced by alkylation treatment, and some metal ion addition introduced by mass spectrometry analysis, such as sodium ions (+21.9819Da, 44271PSMs). Although this experiment has tried to avoid contact with oxygen during sample processing by using a nitrogen operating box and other systems during protein pretreatment, and tried to avoid heating operations, it still introduced a large number of oxidative modifications on methionine (+15.9949Da, 895302PSMs, Figure 6 ); in addition, propionaldehyde (+40.0313Da) was found. This modification mainly occurred on the glycine at the N-terminus, which is consistent with the mass shift of glycine to proline substitution; this modification may be caused by the addition of acetone during the pretreatment process. In addition, there are more deamination oxidation (+0.9840Da) modifications. In addition, the metal ion addition modifications are mainly sodium ions (+21.9819Da), trivalent iron ions (+53.9193Da) and divalent iron ions (+52.9115Da), and calcium ions (+37.9469Da), which mainly occur on the two acidic amino acids of glutamic acid and aspartic acid and the C-terminus of the peptide. Since these sites can generate carboxyl groups with negative charges, they can be added with metal ions with positive charges. The dehydration (-18.0105Da) modification caused by dehydration reaction mainly occurs on tyrosine, threonine and serine containing hydroxyl groups. The above results illustrate the reliability of the modification sites identified by OpenDelta used in this study.

[0211] Proteomics-based analysis of anti-programmed death protein-1 monoclonal antibody therapy for molecular subtypes of non-small cell lung cancer

[0212] First, rank sum analysis (P<0.05) and difference fold (difference fold>2 or difference fold<0.5) were used to obtain 3549 differentially modified sites (866 proteins) screened in peripheral blood mononuclear cells of non-small cell lung cancer patients in the response group and non-response group before anti-programmed death protein-1 monoclonal antibody treatment, of which 102 modified sites (67 proteins) were upregulated and 3447 modified sites (847 proteins) were downregulated; after anti-programmed death protein-1 monoclonal antibody treatment, 2896 differentially modified sites (796 proteins) were screened in peripheral blood mononuclear cells of non-small cell lung cancer patients in the response group and non-response group, of which 435 modified sites (235 proteins) were upregulated and 2461 modified sites (682 proteins) were downregulated. Compared with before anti-PD-1 mAb treatment, a total of 478 differential site modifications (300 proteins) were screened in peripheral blood mononuclear cells of non-small cell lung cancer patients in the response group before and after treatment, of which 206 modification sites (148 proteins) were upregulated and 272 modification sites (182 proteins) were downregulated. After anti-PD-1 mAb treatment, a total of 510 biological modifications were identified in peripheral blood mononuclear cells of non-responding non-small cell lung cancer patients at 33 894 sites of 4196 proteins. Compared with before anti-PD-1 mAb treatment, a total of 78 differential site modifications (63 proteins) were screened in peripheral blood mononuclear cells of non-responding non-small cell lung cancer patients, of which 37 modification sites (44 proteins) were upregulated and 34 modification sites (30 proteins) were downregulated.

[0213] Proteomics-based analysis of anti-programmed death protein-1 monoclonal antibody therapy for molecular subtypes of non-small cell lung cancer

[0214] After summarizing the differential modifications obtained from the above analysis, four feature modules were obtained using the K-means clustering algorithm: Figure 7 ); combined with clinical data, the four modules can be classified into three subtypes according to their modified expression trends: "prediction group", "resistance group" and "benefit group". In summary, according to the expression characteristics of protein modifications before and after anti-programmed death protein-1 monoclonal antibody treatment in responding and non-responding patients, they can be divided into "prediction group" (modules B and C), "benefit group" (modules C and D) and "resistance group" (module A).

[0215] Module B contains 345 biological modifications distributed on 1796 sites of 552 proteins. The expression characteristics of these modifications in peripheral blood mononuclear cells of non-small cell lung cancer before and after treatment with anti-programmed death protein-1 monoclonal antibody are as follows: Before treatment, the expression level of module B protein modification in peripheral blood mononuclear cells of patients in the response group was significantly lower than that in the non-response group. However, anti-programmed death protein-1 monoclonal antibody treatment had no obvious regulatory effect on the protein modification of this module. Therefore, the protein modification of module B can be used to predict whether patients can benefit from anti-programmed death protein-1 monoclonal antibody treatment. Module C contains 445 biological modifications distributed on 2126 sites of 673 proteins. The expression characteristics of these modifications in peripheral blood mononuclear cells of non-small cell lung cancer before and after treatment with anti-programmed death protein-1 monoclonal antibody are as follows: and Module B shows that before treatment with anti-programmed death protein-1 monoclonal antibody, the expression level of module B protein modification in peripheral blood mononuclear cells of patients in the response group was significantly lower than that in the non-response group. However, anti-PD-1 mAb treatment downregulated the protein modification expression of module C, which was close to the response group. Therefore, it may be related to the benefit of anti-PD-1 mAb treatment.

[0216] Modification panel biomarker screening and validation

[0217] The modifications with significant differences in the "prediction group" were screened out, and the receiver operating characteristic curve was drawn. Combined with the corresponding area under the curve value (AUC>0.6), 93 candidate modified peptides were screened and 62 modified peptides were verified by the new cohort using the parallel reaction monitoring method. Potential biomarkers ( Figure 8 ).

[0218] Mass spectrometry-based proteomics has been widely used in the discovery of biomarkers for disease prognosis and diagnosis. The quantitative proteomics method of parallel reaction monitoring (PRM) focuses on a small target set and can be used to measure most proteins or proteins with low protein detection counts due to dynamic range limitations. It is considered to be one of the most sensitive and specific mass spectrometry-based quantitative methods. Therefore, this chapter uses a new clinical cohort (100 cases) and parallel reaction monitoring technology to verify the potential biomarkers screened based on proteomics technology that can be used to predict whether anti-programmed death protein-1 monoclonal antibody treatment is effective. A joint prediction model should be established based on the verification results.

[0219] In this study, the protein subtypes were obtained by using fuzzy mean clustering based pseudo-time sequence analysis (MFUZZ) and K-means clustering algorithm, which were divided into "prediction group", "benefit group" and "resistance group". The protein interaction network of the "prediction group" group proteins was drawn, and the core modules and proteins were obtained by combining MCODE, and the potential protein markers that can predict whether the anti-programmed death protein-1 monoclonal antibody treatment is effective were further screened by combining random forest, Boruta feature analysis and area under the curve (AUC); and a new clinical cohort combined with parallel response monitoring method was used to verify it.

[0220] The present invention finally obtains a combination of markers that can be used to predict whether anti-PD-1 monoclonal antibody treatment is effective in non-small cell lung cancer patients, including 37 proteins (Table 1) and 62 protein-modified peptides (Table 2).

[0221] Table 1 Potential biomarkers in the proteome that can be used to predict whether anti-PD-1 monoclonal antibody therapy is effective in non-small cell lung cancer

[0222] application protein Combined Indicators RHOG, CYBB, CDC42, RHOA, ROCK1, ARF6, S100A8 and S100A9 Combined Indicators LSM3,SNRPD3 Joint indicators FBP1,PGAM1,PKM Joint indicators DEFA1,MMP8,RETN Combined Indicators ASAH1,GM2A Single indicator ARF6 Single indicator EEF1D Single indicator eif2s3 Single indicator FARSB Single indicator GCA Single indicator GNG2 Single indicator H4C9 Single indicator LAMP1 Single indicator LMAN1 Single indicator MNDA Single indicator NAMPT Single indicator NDUFS3 Single indicator PSMC4 Single indicator PSMD6 Single indicator RAB21 Single indicator RAB35 Single indicator RNASE2

[0223] The method for constructing a model for predicting the therapeutic response to anti-programmed death protein-1 monoclonal antibody provided by the present invention uses a pre-treatment subsample and a post-treatment subsample as a pair of paired samples, then performs protein extraction, reduction alkylation, and enzymatic hydrolysis steps on peripheral blood mononuclear cells of the patient samples before and after the anti-programmed death protein-1 monoclonal antibody treatment to obtain sample peptide segments, and after further processing, obtains a biomarker of the therapeutic response effect of the anti-programmed death protein-1 monoclonal antibody, and then monitors the biomarker to obtain a model for predicting the therapeutic response to anti-programmed death protein-1 monoclonal antibody, which can predict the therapeutic effect of the anti-programmed death protein-1 monoclonal antibody for patients with non-small cell lung cancer.

[0224] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0225] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for constructing a prediction model for anti-programmed death protein-1 monoclonal antibody treatment response, characterized in that: The following steps are involved: Obtaining a sample from a patient with non-small cell lung cancer, wherein the patient is a sample from a patient who has been treated with an anti-programmed death protein-1 monoclonal antibody; Extracting peripheral blood mononuclear cells from the patient sample before and after anti-PD-1 monoclonal antibody treatment, wherein for each patient sample, there is a pair of paired samples of a pre-treatment subsample and a post-treatment subsample; The peripheral blood mononuclear cells of the patient sample before and after anti-PD-1 monoclonal antibody treatment are sequentially subjected to steps including protein extraction, reduction alkylation, and enzymatic hydrolysis to obtain sample peptides; Processing the sample peptide segments to obtain processed sample peptide segments; Performing basic feature analysis and functional annotation on the processed sample peptide data to obtain protein spectra and modification spectra of peripheral blood mononuclear cells of patients treated with anti-programmed death protein-1 monoclonal antibody; Based on the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with the anti-programmed death protein-1 monoclonal antibody, machine learning is used to screen biomarkers of anti-programmed death protein-1 monoclonal antibody treatment response effects; The biomarkers of the anti-programmed death protein-1 monoclonal antibody treatment response effect of the non-small cell lung cancer patient samples are monitored to obtain the anti-programmed death protein-1 monoclonal antibody treatment response prediction model.

2. The method for constructing an anti-PD-1 monoclonal antibody treatment response prediction model according to claim 1, characterized in that: In the step of obtaining a sample from a patient with non-small cell lung cancer, wherein the patient is a sample from a patient who has been treated with an anti-programmed death protein-1 monoclonal antibody, The inclusion criteria for the non-small cell lung cancer patient samples include: patients aged 18 years or older, patients clinically diagnosed with non-small cell lung cancer, patients with complete clinical information, and patients who are treated with anti-programmed death protein-1 monoclonal antibody for the first time; The exclusion criteria for the non-small cell lung cancer patient samples include: the patient died during the treatment with anti-programmed death protein-1 monoclonal antibody, or the patient was in the treatment with anti-programmed death protein-1 monoclonal antibody.

3. The method for constructing an anti-PD-1 monoclonal antibody treatment response prediction model according to claim 1, characterized in that: The patient sample is peripheral blood mononuclear cells before and after anti-programmed death protein-1 monoclonal antibody treatment, and the steps of protein extraction, reduction alkylation, and enzymatic hydrolysis are sequentially performed to obtain the sample peptide fragments. The protein extraction includes the following steps: Take out peripheral blood mononuclear cell samples from the liquid nitrogen tank, where the number of cells is 2×10 6 , thawed in a 37℃ water bath, centrifuged at 1200rpm for 3min, and discarded the cryopreservation solution. Prepare 1× intermediate-acting radioimmunoprecipitation lysis buffer RIPA, wherein the volume ratio of intermediate-acting radioimmunoprecipitation lysis buffer to protease inhibitor is 99:

1. According to the amount of peripheral blood mononuclear cells, the intermediate-effect radioimmunoprecipitation lysis buffer was added, wherein, for every 1×10 6 Add 200 μL of medium-release radioimmunoprecipitation lysis buffer to each cell. Add two homogenate beads for lysis, homogenize once, where the frequency is 65 Hz, the total number of grinding times: 2 times, start 45 seconds and stop 10 seconds, incubate on ice for 10 minutes, repeat twice, Centrifuge at 4°C, 12000 rpm for 10 min and remove the supernatant; Accurately pipette 200 μg of protein solution, add 3 times the volume of ice acetone, and freeze in a -80℃ refrigerator for 1h-2h. After 1 hour, take out, centrifuge at 4℃12000rpm for 10min, discard the upper acetone, add 500μL of 80% by weight ice acetone to the lower protein layer, vortex mix, centrifuge again under the same conditions, and discard the supernatant; Add 200 μL of 50 mM ammonium bicarbonate solution and sonicate for 10 min on ice to redissolve the protein; Add 2 μL of 5 mM dithiothreitol to a final concentration of 0.05 mM, vortex to mix, then centrifuge at 1000 rpm for 1 min, and place in a metal bath at 37°C for 30 min; Add 6 μL of 5 mM iodoacetamide to a final concentration of 0.15 mM, vortex to mix, then centrifuge at 1000 rpm for 1 min, and place in a metal bath at 55°C for 30 min. Trypsin hydrolysis: add 10 μL of 0.5 μg / μL sequencing-grade trypsin to make the mass ratio of protein to trypsin 40:

1. Vortex to mix, then centrifuge at 1000 rpm for 1 min, and digest in a metal bath at 37°C and 900 rpm for 12 h. The next day, add 5 μL of sequencing-grade trypsin and digest again under the same conditions for 2 h. After 2 hours, the mixture was taken out, 2 μL of 0.1% formic acid was added, vortexed for 3 seconds to terminate the reaction, and stored in a -80°C refrigerator to obtain sample peptides.

4. The method for constructing an anti-PD-1 monoclonal antibody treatment response prediction model according to claim 3, characterized in that: The sample peptides are processed to obtain the processed sample peptides, which specifically includes the following steps: Install the octadecyl solid phase extraction column on the solid phase extraction device, and place a centrifuge tube under the column to collect the waste liquid; Activation: Add 500 μL 100% acetonitrile to activate the octadecyl solid phase extraction column, repeat twice; Balance: Add 500 μL 0.1% formic acid to wash away residual acetonitrile, repeat twice; Add sample: add peptide to octadecyl solid phase extraction column, collect the effluent, repeat twice and discard the effluent; Desalting: Wash the column with 500 μL 0.1% formic acid, repeat 2-3 times to remove as much salt as possible; Elution of peptides: Add 200 μL of 70% acetonitrile to elute the peptides, repeat twice, and collect the eluate; After peptide desalting, the peptide concentration was quantified using a peptide concentration determination kit; Turn on the low-temperature freeze dryer, set the pressure to 150kpa, and the cold trap temperature to -56℃ for precooling. After the temperature drops to the predetermined value, place the sample, open the centrifuge and vacuum valve in turn, close the atmospheric valve, and vacuum concentrate and dry the 400μL eluate collected by desalting for 3h; After complete freeze-drying, 50 μL of 0.1% formic acid prepared in mass spectrometry water was added to dissolve the freeze-dried peptides. After sufficient vortexing, the mixture was centrifuged at 12,000 rpm at 4°C for 10 min. 45 μL of the supernatant was injected into high performance liquid preparative chromatography for fractionation to obtain the processed sample peptides.

5. The method for constructing an anti-PD-1 monoclonal antibody treatment response prediction model according to claim 4, characterized in that: The classification of the high performance liquid phase preparative chromatography specifically comprises the following steps: 100× stock solution: dilute mass spectrometry grade concentrated ammonia to a concentration of 2M, take 6.7mL of 2M ammonia and add 43.3mL of mass spectrometry water, adjust the pH to 10.5, prepare 50mL of 100x stock solution and place in a 4℃ refrigerator for later use. Aqueous buffer: Take 485 mL of mass spectrometry water and add 10 mL of chromatography grade acetonitrile, then add 5 mL of the prepared 100× stock solution to make a mixed solution of 500 mL of 97% mass spectrometry water, 2% chromatography grade acetonitrile and 1% 100× stock solution. Organic phase buffer: Take 2 mL of mass spectrometry water and add 196 mL of chromatography grade acetonitrile, then add 2 mL of the prepared 100x stock solution to make a mixed solution of 200 mL of 1% mass spectrometry water, 98% chromatography grade acetonitrile and 1% 100x stock solution. 10% methanol: Take 10 mL of chromatography grade methanol and add 90 mL of mass spectrometry water and mix well. After the required solution is prepared, tighten the bottle cap and ultrasonicate for 15 minutes to remove bubbles in the liquid; Under the conditions of column temperature of 40℃, sample tray temperature of 7℃, and pressure of 4000psi, turn on the power supply, column oven, fraction collector, computer and UV monitor in sequence, replace the newly prepared mobile phase, perform wet priming, flush the injector, clean the injection needle, clean the pump head and UV monitoring in sequence, and collect the absorption peaks at wavelengths of 214nm and 280nm. Take 45 μL of sample and inject it into the chromatographic column through the automatic sample injector, then start gradient elution and collect fractions at the same time. The program of the fraction collector is set to collect one tube per minute, for a total of 56 minutes, and a total of 56 tubes are collected. The first tube is discarded, and the remaining 55 tubes are marked with serial numbers. Finally, the eluate is combined into the final 20 fractions in a certain order, concentrated by centrifugation and lyophilized to obtain the processed sample peptides.

6. The method for constructing an anti-PD-1 monoclonal antibody treatment response prediction model according to claim 1, characterized in that: The basic feature analysis and functional annotation are performed on the processed sample peptides to obtain the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with anti-programmed death protein-1 monoclonal antibody, which specifically includes the following steps: The proteome discovery software Protein Discover 2.4 was used to search the mass spectrometry raw data. The 20 flow fractions of each sample were combined and matched with the human protein library downloaded from the protein database https: / / www.uniprot.org / to obtain the protein identification and quantification results. The search parameters of the database were set as follows: the number of amino acids constituting the peptide segment was set to more than 6, a maximum of 2 missed cleavage sites were allowed, the mass tolerance of the parent ion was set to the range of -10ppm-10ppm, and the mass tolerance of the parent ion fragment ion was set to the range of -20ppm--20ppm; The protein quantitative value is standardized using the formula: protein abundance / (average protein of sample / average protein abundance of all samples) to eliminate the protein quantitative value errors between samples caused by sample addition amount, instrument operation, etc. Based on the results identified by Protein Discover 2.4, after removing the contaminating proteins, the quality of mass spectrometry detection data was evaluated by peptide, protein and parent ion mass tolerance distribution analysis. The identified proteins were then subjected to basic feature analysis, annotation of functional databases including Gene Ontology and Kyoto Encyclopedia of Genes and Genomes database, and quantitative analysis, including overall differential analysis of identified proteins, screening of differentially expressed proteins, and expression pattern clustering analysis. The open search engine OpenDelta was used to match the 20 raw files obtained by mass spectrometry for each sample with the human protein library 2022 downloaded from the uniprot database https: / / www.uniprot.org / , and the existing modification database (https: / / www.unimod.org / ) was used for systematic annotation of known modifications, in which acetylation of the protein nitrogen terminus and oxidation on methionine were set as variable modifications, and an open search for known and unknown modifications was performed. The molecular weight deviation range of the precursor ion was set at -10ppm-10ppm, and the molecular weight deviation range of the fragment ion was set at -20ppm-20ppm; The analysis results of the open search engine OpenDelta use (the number of peptide matching spectra of a certain modification at a certain protein site) / (the total number of peptide matching spectra of the unmodified, certain modification, and other arbitrary modifications at the protein site) in order to remove the influence of protein abundance; Based on the identification results of the open search engine, in vitro modifications, including alkylation modifications introduced by pretreatment and oxidation modifications of methionine, were eliminated and potential biomarkers were screened.

7. The method for constructing an anti-PD-1 monoclonal antibody treatment response prediction model according to claim 1, characterized in that: In the step of screening biomarkers of anti-programmed death protein-1 monoclonal antibody treatment response effects based on the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with the anti-programmed death protein-1 monoclonal antibody using machine learning, the screening conditions include: The quantitative values ​​of each protein in two compared sample groups were subjected to T test, and the corresponding P value was calculated as a significance index, and the difference fold > 2 and P value < 0.05 were used as the conditions for screening differentially expressed proteins; The quantitative values ​​of each modification in two compared sample groups were subjected to a rank sum test, and the corresponding P value was calculated as a significance indicator. The P value was corrected using the false discovery rate method, and the difference fold > 2 and P value < 0.05 were used as the conditions for screening differential modifications; The random forest method was used for simulation screening to further explore candidate biomarkers that can be used to predict the benefits of anti-programmed death protein-1 monoclonal antibody treatment. In the random forest analysis, 50,000 trees were constructed, and the maximum number of terminal node trees in the forest was set to 5. In order to evaluate the effectiveness of the model, the subject survival characteristic curve was plotted, and the area under the curve (AUC) of its 95% confidence interval was calculated.

8. The method for constructing an anti-PD-1 monoclonal antibody therapeutic response prediction model according to claim 1, characterized in that: In the step of screening biomarkers of anti-programmed death protein-1 monoclonal antibody treatment response effects based on the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with the anti-programmed death protein-1 monoclonal antibody using machine learning, the analysis method includes: Cluster analysis of protein expression patterns: cluster analysis of differentially expressed proteins was performed using heat maps and pseudo-time sequence analysis based on fuzzy mean clustering (MFUZZ). Gene Ontology and Kyoto Encyclopedia of Genes and Genomes functional enrichment analysis: The internal plug-in ClueGO of Cytoscape3.8.1 software was used to perform gene ontology and Kyoto Encyclopedia of Genes and Genomes functional analysis and enrichment analysis on differentially expressed proteins, mainly including biological pathways and pathway enrichment analysis; Protein interaction network analysis: The target protein was imported into the STRING database to obtain the protein interaction network and the results were imported into Cytoscape 3.8.1 software to draw the protein interaction network diagram. The Cytoscape internal plug-ins Mcode and cytoHubba were used to obtain the corresponding core modules and core nodes respectively.

9. The method for constructing an anti-PD-1 monoclonal antibody therapeutic response prediction model according to claim 1, characterized in that: Based on the protein spectrum and modification spectrum of peripheral blood mononuclear cells of patients treated with the anti-programmed death protein-1 monoclonal antibody, the verification method includes the following steps in which machine learning is used to screen for biomarkers of anti-programmed death protein-1 monoclonal antibody treatment response effects: The protein pretreatment of the validation cohort included protein extraction, reduction, alkylation, enzymatic hydrolysis, and desalting; Proteome: Select peptides of target proteins according to the following criteria to establish parallel reaction monitoring (PRM) method: 1) Only select peptides unique to proteins; 2) No miscleavage sites in peptides; 3) No variable modification sites in peptides, such as oxidation and alkylation; 4) Peptides that meet the above requirements are selected with high response values; 5) At least one and up to three peptides unique to proteins are selected for each protein; Modification group: Parallel reaction monitoring method is established for target modified peptides: 1) There are no mis-cleavage sites in the peptides; 2) There are no variable modification sites in the peptides, such as oxidation modification and alkylation modification; 3) Peptides that meet the above requirements are selected with high response values; The samples were analyzed using a reversed-phase octadecyl self-packed capillary column 75 μm×100 mm and separated using an EASY-nLC 1200 ultra-high performance liquid phase system, wherein mobile phase A was an aqueous solution containing 0.1% formic acid and 2% acetonitrile; mobile phase B was an aqueous solution containing 0.1% formic acid and 90% acetonitrile; the gradient was set as follows: 0-46 min, 6% to 28% B phase; 46-53 min, 28%-45% B phase; 53-60 min, 100% B phase, the flow rate was maintained at 350 nL / min, the column temperature was 50°C, and the peptides were separated by the ultra-high performance liquid phase system and injected into the sodium spray ion source voltage of 2.1 kV, using the ThermoScientificTM Q Exactive HF-X Thermo Fisher mass spectrometer detection, wherein the mass spectrometry data were acquired in a high-sensitivity parallel reaction monitoring mode with the following parameters: high-energy collision dissociation, collision energy of 32%, primary resolution of 60,000, secondary resolution of 15,000, scanning range of 100-1800 m / z, scanning time of 100 ms, screening of charge states of +2 to +4 including precursors, dynamic exclusion time of 30 s, isolation window of 4 Da, and time window of 7 min; Skyline 3.6 software was used for parallel reaction monitoring data processing. All results were imported into the software, and the correct peaks were selected to obtain the results of all peptides in all samples. Among them, the parallel reaction monitoring results, including protein names and peak areas, were further exported for analysis to screen and compare differential proteins in different groups; Based on the relative quantitative results obtained by the Skyline peak area, the R language was used for correlation analysis and visualization of the differences, and the R package was used to draw the receiver operating characteristic curve and calculate the area under the curve.

10. The method for constructing a prediction model for anti-PD-1 monoclonal antibody treatment response according to claim 1, characterized in that: In the step of monitoring the biomarkers of the anti-programmed death protein-1 monoclonal antibody treatment response effect of the non-small cell lung cancer patient samples and obtaining the anti-programmed death protein-1 monoclonal antibody treatment response prediction model, the non-small cell lung cancer patient samples are classified into three subtypes: prediction group, resistance group, and benefit group according to the expression trends of modular proteins and modifications.

Images