Myotrophic lateral sclerosis prediction and prognosis evaluation system

Abstract

The invention provides an amyotrophic lateral sclerosis prediction and prognosis evaluation system and a construction method of a sALS prediction and prognosis evaluation model based on multiple proteins. The system comprises an input module, a prediction evaluation module and an output module. The input module is used for acquiring the content of multiple candidate ALS serum biomarkers. And the prediction and evaluation module is used for calculating and scoring according to the sALS prediction and prognosis evaluation model based on multiple proteins, judging that the patient suffers from amyotrophic lateral sclerosis when the scoring result is greater than or equal to a specific judgment threshold value, and judging that the patient does not suffer from amyotrophic lateral sclerosis when the scoring result is less than the specific judgment threshold value. And the output module is used for outputting the prediction evaluation result. After the technical scheme is adopted, amyotrophic lateral sclerosis can be simply, quickly and accurately predicted and evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of disease prediction, and in particular to an amyotrophic lateral sclerosis prediction and prognosis evaluation system. Background Art

[0002] Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease that can lead to severe disability and death. At present, the pathogenic mechanism of ALS is still unclear, and there is no effective therapeutic target or satisfactory treatment plan. The prevalence of ALS is 5 / 100,000, the incidence is 1.7 / 100,000, and the average survival period is short (median survival period is 2-5 years). Although ALS was first described in 1869, its diagnosis still mainly relies on medical history and neurological examination. However, when characteristic clinical symptoms or electrophysiological changes appear, ALS has progressed to the middle stage of neurodegeneration, a large number of motor neurons have been lost, and the course of the disease is difficult to reverse. Therefore, early diagnosis and early intervention of ALS are particularly important. At present, the diagnosis of ALS is mainly based on the patient's clinical manifestations and the doctor's clinical experience, which has led to many patients being missed or misdiagnosed. Therefore, on the one hand, it is urgent to find sensitive and reliable biological markers, and on the other hand, simple and economical auxiliary examination methods are needed to help the diagnosis of ALS. In addition, ALS is a progressive disease. If ALS biomarkers can be further used to monitor disease progression and predict patient prognosis, it will be of great benefit to the clinical diagnosis and treatment of ALS.

[0003] There are two forms of ALS: familial and sporadic ALS (fALS and sALS). sALS accounts for 90% of ALS patients, while fALS accounts for the remaining 10%. fALS is diagnosed through gene mutation testing and family history, while sALS is difficult to diagnose early. The cause of sALS is still unclear, and several different pathogenic mechanisms have been proposed. There is an urgent need to identify biomarkers for sALS.

[0004] Existing research on ALS biological markers mainly focuses on clinical symptoms, electrophysiology, neuroimaging and biochemistry:

[0005] 1. Clinical symptoms of ALS: Motor symptoms are currently the main basis for the diagnosis of ALS. However, upper and lower motor neuron symptoms only appear after significant neuronal damage occurs and cannot be used as a marker for early diagnosis. Non-motor symptoms such as behavioral changes and executive dysfunction may appear in some patients earlier than motor symptoms, but their sensitivity and specificity for the diagnosis of ALS are not high.

[0006] 2. Neuroelectrophysiology: Concentric needle electromyography is an important auxiliary examination for the diagnosis and differential diagnosis of ALS. Extensive active denervation and chronic renervation can detect lower motor neuron lesions earlier than physical examination. However, similar to clinical symptoms, when patients show significant electromyographic changes, they are often in the middle and late stages of the disease, and the specificity and sensitivity of electromyography for early patients are not ideal. Magnetic stimulation motor evoked potentials can help detect upper motor neuron lesions in ALS, but the sensitivity is not high.

[0007] 3. Neuroimaging: Conventional imaging examinations such as magnetic resonance imaging are difficult to provide a basis for the diagnosis of ALS. They are mainly used to assist in differential diagnosis and exclude structural damage. Functional magnetic resonance imaging, motor cortex thickness analysis, magnetic resonance spectroscopy, pyramidal beam diffusion tensor imaging, positron emission tomography, single photon emission computed tomography and other technologies can be used as biomarkers to reflect the manifestations of upper motor neuron involvement, but are still in the research stage and are difficult to promote due to their high prices, poor practicality and poor specificity.

[0008] 4. Biochemistry: Research on biochemical markers for early diagnosis of ALS involves multiple fields such as immunity, inflammation, oxidative stress, and apoptosis, among which neurofilament light chain protein has the greatest potential auxiliary diagnostic value. Increased levels of neurofilament light chain protein in cerebrospinal fluid and serum can indicate upper motor neuron lesions in ALS, but this marker lacks specificity and is difficult to be directly applied to the early diagnosis and differential diagnosis of ALS.

[0009] The diagnosis of sALS is still based on clinical symptoms and neurological and electrophysiological examinations. Due to the large clinical heterogeneity and the lack of effective biomarkers, once patients meet the diagnostic criteria for sALS, they usually have lost most motor neurons, limiting the therapeutic potential of various potential therapies. In addition, there are still difficulties in assessing disease progression and treatment efficacy in sALS. Some quantitative or semi-quantitative functional scales such as ALSFRS-R scores have been used to assess disease progression and the effectiveness of treatment. However, due to low sensitivity, the utility of these scales remains controversial. Therefore, a biological marker for the diagnosis of ALS with high sensitivity and specificity that can be widely used in clinical practice has not yet been found. Identifying candidate biomarkers can promote early diagnosis, early treatment, accurate disease progression assessment and treatment efficacy.

[0010] Cerebrospinal fluid is the most direct specimen that reflects pathological changes in the central nervous system. However, the collection of cerebrospinal fluid is highly invasive and requires high-risk operations, and it is difficult to obtain patient consent for repeated sample collection for disease monitoring; in contrast, blood tests are invasive, simple to operate, and inexpensive, making them more suitable for disease screening. Previously reported blood markers that may be helpful in the diagnosis of ALS include inflammatory factors, neurofilament light chains of plasma proteins, and ubiquitin C-terminal hydrolase L1. This suggests that blood tests can reflect the disease characteristics of ALS to a certain extent, and blood biochemical markers have the potential to help diagnose and prognose ALS. However, due to the lack of verification and small sample size of the above studies, the above potential ALS biomarkers are still controversial.

[0011] In recent years, proteomics has been increasingly used to explore biomarkers and describe the full picture of the disease, helping to better understand the disease mechanism. Proteomic analysis has the advantages of high throughput, objectivity and quantification, helping to understand multifactorial pathophysiological processes and contribute to the development of effective therapeutic interventions. Some studies have used high-throughput proteomics methods to explore new ALS biomarkers. However, due to methodological differences and small sample sizes, the reproducibility of these studies is limited. In addition, focusing on a single pathway or a specific type of functional protein limits the ability to describe the full picture of the disease. Previous reports on ALS biomarkers have mainly focused on a single protein or pathway. However, the diagnostic efficiency of a single protein is limited, which may be because sALS is a complex disease with multiple pathological mechanisms, and a single protein or pathway cannot fully reflect a highly complex disease. Summary of the invention

[0012] In order to overcome the above technical defects, the first aspect of the present invention provides a method for constructing an amyotrophic lateral sclerosis prediction and prognosis evaluation model, which comprises:

[0013] Step S1: Based on the test set consisting of ALS patient data and healthy control group data, several potential ALS serum biomarkers were preliminarily screened by label-free quantitative proteomics;

[0014] Step S2: Based on the validation set consisting of ALS patient data and healthy control group data, the expression levels of the potential ALS serum biomarkers are detected by ELISA detection kit, and then the ALS serum biomarkers with significantly upregulated expression levels are sequentially subjected to principal component dimensionality reduction analysis and interaction analysis, so as to finally screen out several ALS serum biomarkers, and perform receiver operating characteristic analysis to evaluate the sensitivity and specificity of each biomarker in distinguishing sALS patients from healthy controls;

[0015] Step S3: Based on two or more of the several ALS serum biomarkers, a logistic regression model is constructed by using the likelihood ratio method to generate a sALS prediction and prognosis evaluation model based on multiple proteins, wherein the sALS prediction and prognosis evaluation model based on multiple proteins is used to comprehensively calculate a scoring result according to the concentrations of multiple proteins. If the scoring result is greater than or equal to a specific judgment threshold, it is judged that the patient suffers from amyotrophic lateral sclerosis.

[0016] Furthermore, in step S1, false discovery rate analysis was used to identify differentially expressed proteins; principal multivariate component analysis, gene ontology enrichment analysis, pathway enrichment analysis, and protein interaction network analysis were performed using the STRING database to determine the functional clustering of differentially expressed proteins; and several potential ALS serum biomarkers were further screened out based on the degree of expression difference, molecular function, and organ expression specificity.

[0017] Furthermore, the several candidate ALS serum biomarkers include proteins encoded by FKBP1A, CD84, CAMP, ZYX, HBA1, HBB, TLN1, and TPT1 genes, respectively.

[0018] Furthermore, the sALS prediction and prognosis assessment model based on multiple proteins is:

[0019] Log(P)=0.002*HBB+0.762*FKBP1A+0.183*CD84+0.589*CAMP+0.172*TLN1+

[0020] 0.744*ZYX+0.047*HBA1+0.209*TPT1-8.476

[0021] Wherein, HBB in the formula represents the protein concentration encoded by the HBB gene, FKBP1A represents the protein concentration encoded by the FKBP1A gene, CD84 represents the protein concentration encoded by the CD84 gene, CAMP represents the protein concentration encoded by the CAMP gene, TLN1 represents the protein concentration encoded by the TLN1 gene, ZYX represents the protein concentration encoded by the ZYX gene, HBA1 represents the protein concentration encoded by the HBA1 gene, and TPT1 represents the protein concentration encoded by the TPT1 gene;

[0022] The judgment threshold of the sALS prediction and prognosis evaluation model based on multiple proteins is 0.528.

[0023] Furthermore, the sALS prediction and prognosis assessment model based on multiple proteins is:

[0024] Log(P)=0.238*TLN1+0.175*TPT1+1.446*ZYX-5.644

[0025] Wherein, TLN1 in the formula represents the concentration of protein encoded by the TLN1 gene, TPT1 represents the concentration of protein encoded by the TPT1 gene, and ZYX represents the concentration of protein encoded by the ZYX gene;

[0026] The judgment threshold corresponding to the sALS prediction and prognosis evaluation model based on multiple proteins is 0.351;

[0027] The model is used for the prediction of early-stage amyotrophic lateral sclerosis.

[0028] The second aspect of the present application provides an amyotrophic lateral sclerosis prediction and prognosis assessment system, comprising:

[0029] An input module, the input module is used to obtain the content of multiple candidate ALS serum biomarkers;

[0030] A prediction and evaluation module, wherein the prediction and evaluation module is used to calculate a score based on a sALS prediction and prognosis evaluation model based on multiple proteins, and when the score result is greater than or equal to a specific judgment threshold, it is judged that the patient suffers from amyotrophic lateral sclerosis, and when the score result is less than the specific judgment threshold, it is judged that the patient does not suffer from amyotrophic lateral sclerosis;

[0031] An output module is used to output the prediction evaluation result.

[0032] Furthermore, the several candidate ALS serum biomarkers are selected from two or more of the proteins encoded by FKBP1A, CD84, CAMP, ZYX, HBA1, HBB, TLN1, and TPT1 genes.

[0033] Furthermore, the sALS prediction and prognosis assessment model based on multiple proteins is:

[0034] Log(P)=0.002*HBB+0.762*FKBP1A+0.183*CD84+0.589*CAMP+0.172*TLN1+

[0035] 0.744*ZYX+0.047*HBA1+0.209*TPT1-8.476

[0036] Wherein, HBB in the formula represents the protein concentration encoded by the HBB gene, FKBP1A represents the protein concentration encoded by the FKBP1A gene, CD84 represents the protein concentration encoded by the CD84 gene, CAMP represents the protein concentration encoded by the CAMP gene, TLN1 represents the protein concentration encoded by the TLN1 gene, ZYX represents the protein concentration encoded by the ZYX gene, HBA1 represents the protein concentration encoded by the HBA1 gene, and TPT1 represents the protein concentration encoded by the TPT1 gene;

[0037] The judgment threshold of the sALS prediction and prognosis evaluation model based on multiple proteins is 0.528.

[0038] Furthermore, the sALS prediction and prognosis assessment model based on multiple proteins is:

[0039] Log(P)=0.238*TLN1+0.175*TPT1+1.446*ZYX-5.644

[0040] Wherein, TLN1 in the formula represents the concentration of protein encoded by TLN1 gene, TPT1 represents the concentration of protein encoded by TPT1 gene, and ZYX represents the concentration of protein encoded by ZYX gene;

[0041] The judgment threshold corresponding to the sALS prediction and prognosis evaluation model based on multiple proteins is 0.351;

[0042] The model is used for the prediction of early-stage amyotrophic lateral sclerosis.

[0043] Compared with the prior art, the above technical solution has the following beneficial effects:

[0044] This application uses proteomics to perform initial screening of potential sALS biomarkers, and further validates them in a larger sample size of patients, ultimately identifying a more reliable combination of multiple biomarkers, and on the basis of the combination of multiple biomarkers, establishing a sALS prediction and prognostic evaluation model based on multiple proteins, thereby achieving rapid quantitative prediction and evaluation of the early and prognostic diagnosis of sALS. For example, (1) a logic model based on eight proteins (HBB, FKBP1A, CD84, CAMP, TLN1, ZYX, HBA1, and TPT1) showed significant effectiveness in distinguishing sALS from the control group (AUC: 0.833, P < 0.0001). (2) Since clinical symptoms are not obvious in the early stages of sALS, neurological and electrophysiological examinations may not be able to identify it. In order to diagnose early (onset < 6 months of disease course) sALS, we developed a logic model group based on three proteins (TLN1, TPT1, and ZYX), which is more helpful for early diagnosis. (3) sALS patients with lower ALSFRS-R scores showed higher expression of three proteins (FKBP1A, CAMP, and HBA1) than those with higher ALSFRS-R scores, which has significant potential in monitoring disease prognosis when combined with the assessment of clinical parameters.

[0045] In addition, our proteomic analysis provides new insights into the pathogenesis of ALS, which can help in the diagnosis of ALS and detection of disease progression, and improve risk prediction techniques and screening of intervention targets. Because sALS is a complex disease with multiple pathological mechanisms, a single protein or pathway cannot fully reflect a highly complex disease. Therefore, we proposed a panel of proteins including those from multiple pathways to help distinguish sALS patients from controls. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 There were differences in the expression of eight candidate biomarkers between sALS and normal controls (*P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, statistical method was one-way analysis of variance with the smallest difference after the test);

[0047] Figure 2 Multivariate logistic regression analysis of eight candidate biomarkers suggested that CAMP had the best diagnostic efficacy for sALS;

[0048] Figure 3 is the diagnostic efficacy curve for sALS based on eight biomarkers;

[0049] Figure 4It is a combination of markers for early diagnosis and prognosis monitoring of sALS. a and b are multivariate logistic regression analysis based on the combination of three biomarkers, indicating that it has better diagnostic efficacy for early sALS. c is the difference in protein expression levels determined by ELISA and its correlation with different ALSFRS-R scores (*P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, the statistical method is one-way analysis of variance with the smallest difference after test). DETAILED DESCRIPTION

[0050] The advantages of the present invention are further described below in conjunction with the accompanying drawings and specific embodiments. Those skilled in the art should understand that the following specific description is illustrative rather than restrictive, and should not be used to limit the scope of protection of the present invention.

[0051] This embodiment provides a process for constructing a sALS prediction and prognosis evaluation model based on multiple proteins, a process for constructing an amyotrophic lateral sclerosis prediction and prognosis evaluation system, and applications thereof.

[0052] 1. Construction of a sALS prediction and prognostic assessment model based on multiple proteins

[0053] In order to better develop a diagnostic and prognostic system for ALS, we used proteomics to perform an initial screening of potential biomarkers, and further validated them in a larger sample size of patients, ultimately identifying a more reliable biomarker combination to assist in the diagnosis and prognosis of ALS.

[0054] The method for constructing a prediction and prognosis assessment model for amyotrophic lateral sclerosis includes steps S1 to S3:

[0055] Step S1: Based on the test set consisting of ALS patient data and healthy control group data, several potential ALS serum biomarkers were preliminarily screened by label-free quantitative proteomics.

[0056] We recruited 10 sporadic ALS patients and 5 healthy controls to screen potential ALS serum biomarkers by label-free quantitative proteomics. False discovery rate (FDR) analysis was used to identify differentially expressed proteins (DEPs); principal component analysis (PCA), gene ontology (GO) enrichment analysis, pathway enrichment analysis (KEGG), and protein interaction network analysis (PPI) were performed using the STRING database to determine the functional clustering of DEPs. Candidate biomarkers were further selected based on the degree of expression difference, molecular function, and organ expression specificity.

[0057] Step S2: Based on the validation set consisting of ALS patient data and healthy control group data, the expression levels of the potential ALS serum biomarkers were detected by ELISA detection kit, and then the ALS serum biomarkers with significantly increased expression levels were subjected to principal component dimensionality reduction analysis and interaction analysis in turn, so as to finally screen out several ALS serum biomarkers, and perform receiver operating characteristic analysis to evaluate the sensitivity and specificity of each biomarker in distinguishing sALS patients from healthy controls. These biomarkers have the potential to be used for early diagnosis of sALS, disease progression monitoring, and quantitative evaluation of treatment effects, which are crucial in clinical practice.

[0058] The diagnostic value of the candidate biomarkers was verified by detecting the above serum markers by ELISA in a validation cohort of 125 sporadic ALS patients and 125 controls, and a multi-protein combination kit was further established to improve the diagnostic efficacy, achieve early diagnosis and evaluate disease progression.

[0059] 1682 proteins were detected by label-free quantitative proteomics analysis, of which 387 were identified as DEPs associated with sALS. Among these 387 proteins, 259 proteins were upregulated and 128 proteins were downregulated in sALS patients compared with controls. Principal multivariate component analysis showed that DEPs separated sALS patients from controls. In addition, cluster heat map analysis generated according to the overall DEP expression trend could distinguish sALS patients. According to GO analysis, the molecular functions of DEPs were mainly related to antioxidants, protein disulfide reductase, carbonyl lyase, hydrolase, and peroxidase activities. These results suggest that antioxidant stress plays an important role in the pathogenesis of ALS, which is consistent with previous studies that oxidative stress damage is involved in the pathogenesis of ALS. KEGG pathway analysis showed that DEPs were enriched in platelet activation, energy metabolism (glycolysis, tricarboxylic acid cycle, carbon metabolism, amino acid biosynthesis), and peroxisomes. The DEPs significantly upregulated in ALS were subjected to principal component dimensionality reduction analysis, with FDR < 0.05 and expression levels differing by more than 5 times, generating 42 candidate biomarker proteins. Interaction analysis was performed on the 42 proteins to screen for biomarkers. Proteins representing different pathways were selected as much as possible and expressed in the central nervous system, which narrowed the scope to 8 DEPs (encoded by genes FKBP1A, CD84, CAMP, ZYX, HBA1, HBB, TLN1, and TPT1, respectively), which were considered strong candidate biomarkers and further validated in the validation cohort.

[0060] The performance of eight potential diagnostic biomarkers was validated in a cohort of 125 sALS patients and 125 healthy controls. These biomarkers have the potential to be used for early diagnosis of sALS, monitoring of disease progression, and quantitative assessment of treatment effects, which are crucial in clinical practice. The expression levels of all eight potential biomarkers in sALS patients were significantly higher than those in the control group (see Appendix Figure 1 ), which is upregulated more than 5 times in the central nervous system of sALS patients, and they are involved in multiple functional pathways. Receiver operating characteristic analysis was performed to evaluate the sensitivity and specificity of each biomarker in distinguishing sALS patients from controls, and it was found that cathelicidin-associated antimicrobial peptide (CAMP) protein had the strongest ability to distinguish between sALS and controls (AUC: 0.711, P < 0.0001) (see Appendix Figure 2 ).

[0061] Step S3: Based on two or more of the several ALS serum biomarkers, a logistic regression model is constructed by using the likelihood ratio method to generate a sALS prediction and prognosis evaluation model based on multiple proteins, wherein the sALS prediction and prognosis evaluation model based on multiple proteins is used to comprehensively calculate a scoring result according to the concentrations of multiple proteins. If the scoring result is greater than or equal to a specific judgment threshold, it is judged that the patient suffers from amyotrophic lateral sclerosis.

[0062] Due to the limited diagnostic efficiency of a single protein, we next developed a multi-protein diagnostic panel by constructing a logistic regression model using the likelihood ratio method. Ultimately, a protein combination system containing eight markers (HBB, FKBP1A, CD84, CAMP, TLN1, ZYX, HBA1, and TPT1) with high discrimination ability was obtained (AUC: 0.833, P < 0.0001). The logistic model based on the eight proteins generated a dichotomous score that can be used to accurately characterize each sample. The probability score P of a sample with a positive diagnosis of sALS for each protein marker value is defined as:

[0063] (1) By way of example, in a preferred embodiment, the sALS prediction and prognosis assessment model based on multiple proteins is:

[0064] Log(P)=0.002*HBB+0.762*FKBP1A+0.183*CD84+0.589*CAMP+0.172*TLN1+

[0065] 0.744*ZYX+0.047*HBA1+0.209*TPT1-8.476

[0066] Among them, HBB in the formula represents the protein concentration encoded by the HBB gene, FKBP1A represents the protein concentration encoded by the FKBP1A gene, CD84 represents the protein concentration encoded by the CD84 gene, CAMP represents the protein concentration encoded by the CAMP gene, TLN1 represents the protein concentration encoded by the TLN1 gene, ZYX represents the protein concentration encoded by the ZYX gene, HBA1 represents the protein concentration encoded by the HBA1 gene, and TPT1 represents the protein concentration encoded by the TPT1 gene.

[0067] The probability score of the sALS group was significantly higher than that of the control group. The judgment threshold of the sALS prediction and prognosis evaluation model based on eight proteins was 0.528. The sensitivity and specificity scores for diagnosing sALS at the threshold point of 0.528 were 76% and 76.8%, respectively, with an AUC of 0.833, P < 0.0001 (see Appendix Figure 3 ).

[0068] To further test the clinical application effect of this model, the model was applied to a new cohort (25 sALS patients and 25 normal controls) to calculate Log(P). In the sALS group, 24 patients had Log(P)>0.528, and 1 patient had Log(P)<0.528; in the normal control group, 21 patients had Log(P)<0.528, and 4 patients had Log(P)>0.528 (Fisher's exact test P<0.0001). The diagnostic sensitivity of this model was 96%, the specificity was 84%, the positive predictive value was 85.71%, and the negative predictive value was 95.45%, which showed good diagnostic efficacy.

[0069] The prediction model based on eight proteins (HBB, FKBP1A, CD84, CAMP, TLN1, ZYX, HBA1, and TPT1) is not only applicable to the early prediction of sALS (onset < 6 months of disease course) but also to prognostic assessment.

[0070] (2) By way of example, in another preferred embodiment, the sALS prediction and prognosis assessment model based on multiple proteins is:

[0071] Log(P)=0.238*TLN1+0.175*TPT1+1.446*ZYX-5.644

[0072] Among them, TLN1 in the formula represents the protein concentration encoded by the TLN1 gene, TPT1 represents the protein concentration encoded by the TPT1 gene, and ZYX represents the protein concentration encoded by the ZYX gene. This model is particularly suitable for the prediction of early amyotrophic lateral sclerosis.

[0073] The logistic regression model based on the combination of three proteins (TLN1, TPT1 and ZYX) had an AUC of 0.784 in early sALS (onset < 6 months of disease duration), which is helpful for the early diagnosis of sALS (see Appendix Figure 4 ). The judgment threshold of the sALS prediction and prognosis evaluation model based on three proteins was 0.351. The probability score of early sALS was significantly higher than that of the control group. At the threshold point of 0.351, the sensitivity and specificity scores for diagnosing early sALS were 82% and 68%, respectively.

[0074] (3) In addition, the expression levels of three proteins (FKBP1A, CAMP, and HBA1) were significantly different between patients with different ALSFRS-R levels (low, medium, and high), and higher protein expression was associated with lower ALSFRS-R scores (see Appendix Figure 4 ), which helps in prognosis. sALS patients with lower ALSFRS-R scores showed higher expression of three proteins (FKBP1A, CAMP and HBA1) than those with higher ALSFRS-R scores, suggesting that these three proteins may be helpful in monitoring disease progression. FKBP1A, CAMP and HBA1 have significant potential in monitoring disease progression in combination with the assessment of clinical parameters.

[0075] The details to implement the above build steps are as follows:

[0076] 1. Patient enrollment

[0077] From January 2020 to December 2021, 160 sALS patients were recruited from the Department of Neurology and Research Institute of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine. All patients were recruited according to the revised El Escorial diagnostic criteria to determine ALS. The enrolled patients were excluded from having other diseases affecting the peripheral nerves, including diabetes. During the recruitment process, medical history, physical examination, and biochemical analysis were collected. The ALS Functional Rating Scale-Revised (ALSFRS-R) was used to assess the severity of the disease. The ALSFRS-R scale has 12 items representing four functional subdomains. These include the medulla oblongata, fine motor, gross motor, and respiration, and the score of each item ranges from 0 (complete loss of function) to 4 (no loss of function). The total score ranges from 0 to 48 points; the higher the score, the better the function. At the same time, we recruited 155 age- and sex-matched healthy controls from the physical examination center of Ruijin Hospital.

[0078] Finally, 10 sALS patients and 5 controls were randomly selected as the discovery cohort for quantitative label-free proteomic analysis. In addition, 125 sALS patients and 125 controls were randomly assigned to the validation cohort that received ELISA to determine the target biomarkers identified from the discovery cohort. All 125 sALS patients were divided into early (onset <6 months of disease course, n = 45) and late symptomatic sALS (onset >6 months of disease course, n = 80). In addition, according to the ALSFRS-R score during recruitment, the patients were divided into three groups: low score group (≤25, n = 34), medium score group (26-32, n = 61), and high score group (≥33, n = 30). In addition, another validation cohort consisting of 25 sALS patients and 25 controls was used to test the diagnostic efficacy of the predictive evaluation system. The Human Ethics Committee of Ruijin Hospital approved this study (approval number: 2020-No.50), and all participants signed informed consent.

[0079] 2. Collection of serum samples

[0080] Blood samples were collected by venipuncture in a serum separation vacuum. After standing for 1-2 hours at room temperature, serum samples were aliquoted and stored at -80°C for proteomic analysis after centrifugation (3000 rpm, 10 minutes at 4°C).

[0081] 3. Proteomic analysis

[0082] 3.1 Protein extraction

[0083] Serum samples were centrifuged at 11,000 × g for 10 min at 4°C to remove cell debris. The supernatant was transferred to a new centrifuge tube. The Top 12 High Abundance Protein Removal PierceTM Spin Columns Kit (Thermo Fisher Scientific, Hanover Park, IL, USA) was used to remove high abundance serum proteins to eliminate interference. Protein concentration was determined using the bicinchoninic acid assay (BCA) according to the manufacturer's instructions (Thermo Fisher Scientific).

[0084] 3.2 Trypsin digestion

[0085] After removal of high-abundance serum proteins, samples were dried using a cryo-vacuum concentrator (Thermo Fisher Scientific). 8 M urea was added to re-dissolve the remaining protein, dithiothreitol was added to a final concentration of 5 mM, and the protein was reduced at 56 °C for 30 min. Iodoacetamide was added to a final concentration of 11 mM, and the mixture was incubated at room temperature and protected from light for 15 min, after which the alkylated sample was transferred to an ultrafiltration membrane (10 kDa molecular weight cutoff membrane, Millipore, Darmstadt, Germany) and centrifuged at 12,000 × g for 20 min at room temperature. The sample was resuspended three times in 8 M urea and then three times in ammonium bicarbonate, each time centrifuged, and then trypsin was added at a mass ratio of 1:50 (protease: protein, M / M), and the sample was digested overnight at 37 °C. The supernatant was collected by centrifugation at 12,000 × g for 10 min at room temperature, and purified water was added to increase solubility. The enzymatically digested peptide solution was acidified to pH 2–3 using 10% trifluoroacetic acid and then centrifuged at 12,000 × g for 10 min at room temperature, and the supernatant was transferred to a new centrifuge tube for Stage Tip (Pierce, Thermo Fisher Scientific) desalting.

[0086] 3.3 Liquid chromatography and mass spectrometry

[0087] Tryptic peptides were analyzed using an EASY nLC 1200 ultra-high performance liquid chromatography system (Thermo Fisher Scientific), where mobile phase A was an aqueous solution containing 0.1% formic acid and 2% acetonitrile, and mobile phase B was an aqueous solvent containing 0.1% formic acid and 90% acetonitrile. The liquid phase gradient was set as 0-96 min, 4%-20% B; 96-114 min, 20%-32% B; 114-117 min, 32%-80% B; 117-120 min, 80% B, and the flow rate was maintained at 500.00 nL / min. The peptides were separated by the UHPLC system and then injected into the Nanospray Flex TMElectron source (Thermo Fisher Scientific) was used for ionization. The separated peptides were detected using an Exploris 480 mass spectrometer system (Thermo Fisher Scientific). The ion source voltage was set to 2.2 kV, and a high-resolution Orbitrap was used to detect and analyze the peptide parent ions and their secondary fragments. The primary mass spectrometer scan range was set to 400–1200 m / z with a resolution of 60000.00, while the secondary mass spectrometer scan range was fixed at 100 m / z with a resolving power of 30000.00. The data acquisition mode used a data-dependent acquisition scanning procedure, that is, the first 15.00 peptide parent ions with the highest signal intensity were selected after the primary scan. The fragmentation energy of 27% was used to sequentially enter the high-energy collision-dissociation-collision cell for fragmentation, and the secondary mass spectrometry analysis was performed in sequence. To improve the effective utilization of the mass spectrometer, the automatic gain control was set to 7.5E4, the signal threshold was set to 1E4 ions / s, the maximum injection time was set to 100 ms, and the dynamic exclusion time of the tandem mass spectrometry scan was set to 30 s to avoid repeated scanning of the parent ion.

[0088] 3.4 Database Search

[0089] Proteome DiscoverTM 2.4 (Thermo Fisher Scientific) was used to retrieve the secondary mass spectrometry data. The search parameters were set as follows: the database used was Homo_sapiens_9606 (20366 sequences), a reverse library was added to calculate the false discovery rate caused by random matching, and a common contamination library was added to the database to eliminate the influence of contaminating proteins in the identification results; the enzyme cleavage mode was set to trypsin (Full); the number of missed cleavage sites was set to two; the minimum peptide length was set to 7 amino acid residues; the maximum number of peptide modifications was set to five; the mass error tolerance of the primary parent ion for the first search and the main search was set to 10ppm and 5ppm, respectively, and the mass error tolerance of the second fragment ion was set to 0.02Da. The alkylation aminomethyl of cysteine ​​was set as a fixed modification and a variable modification a'['Acetyl(Protein N-te'm)','Oxidation'M)','Deamidation('Q)']. The quantitative method was set to LFQ, and the false discovery rate of protein identification and peptide spectrum matching identification was set to 1%.

[0090] 3.5 Functional cluster analysis

[0091] GO functional clustering was performed to determine the biological functions of proteins with significantly different expression levels in the serum of sALS patients and healthy controls. Two-tailed Fisher's exact test was used to classify proteins annotated using the GO database. In addition, a two-tailed Fisher's exact test was performed using the KEGG (https: / / www.genome.jp / kegg / ) database to perform pathway analysis to determine enriched pathways. Each protein category was searched using the InterPro database (https: / / www.ebi.ac.uk / interpro) and a two-tailed Fisher's exact test was performed. Functional classification (GO, domain, and pathway) was performed based on differentially expressed proteins (DEPs), and categories with at least one enriched cluster were included in the hierarchical clustering analysis. The function x=-log10(P value) was used to transform the filtered P value matrix. Then, the x value of each category was z-transformed. The z scores were clustered using the one-way hierarchical clustering function of the Genesis software program. For cluster membership visualization, a heatmap was generated using the "heatmap.2" function in the "gplot" tool of the R package software program. The molecular interactions between DEPs were analyzed using the STRING database (version 10.1; https: / / string-db.org)6 and visualized using the “networkD3” tool of the R package software program (version 0.4, https: / / CRAN.R-project.org / package=networkD3).

[0092] 4. Verification using ELISA

[0093] The expression levels of cathelicidin-related antimicrobial peptide (CAMP, ELK5115, ELK Biotechnology, Wuhan, China), FK506 binding protein 1A (FKBP1A, ELK4323, ELK Biotechnology), hemoglobin subunit alpha (HBA1, ELK5194, ELK Biotechnology), hemoglobin subunit beta (HBB, ELK4071, ELK Biotechnology), CD84 (ELK8729, ELK Biotechnology), talin-1 (TLN1, ELK2345, ELK Biotechnology), translationally controlled tumor protein (TPT1, ELK4011, ELK Biotechnology), and zyxin (ZYX, ELK4891, ELK Biotechnology) were detected using the corresponding ELISA detection kits according to the manufacturer's instructions.

[0094] 5. Statistical analysis

[0095] Categorical data were analyzed using chi-square test or Fisher's exact test, and continuous data were evaluated using Student's t test or one-way analysis of variance and post hoc least significant difference test. Multivariate logistic regression analysis was performed to develop a diagnostic algorithm and score for sALS. The area under the receiver operating curve (area under the curve, AUC) was determined and compared using the Z test. A two-tailed P value <0.05 was considered to indicate a significant difference. Statistical analyses were performed using SPSS software (version 18.0, SPSS Inc., Chicago, IL, USA).

[0096] 2. Construction and application of the prediction and prognostic assessment system for amyotrophic lateral sclerosis

[0097] Based on the above-mentioned sALS prediction and prognosis evaluation model based on multiple proteins, a prediction and prognosis evaluation system for amyotrophic lateral sclerosis is constructed by computer program, and the system includes an input module, a prediction evaluation module and an output module. When the computer program is executed, the functions of the input module, the prediction evaluation module and the output module can be realized.

[0098] The input module obtains the contents of multiple candidate ALS serum biomarkers. In practical applications, only the concentration data of specific types of ALS serum biomarkers involved in the sALS prediction and prognosis evaluation model based on multiple proteins need to be input.

[0099] The prediction and evaluation module calculates the score based on the input concentration data of a specific type of ALS serum biomarker and according to the pre-set sALS prediction and prognosis evaluation model. When the score result is greater than or equal to a specific judgment threshold, it is judged that the patient has amyotrophic lateral sclerosis. When the score result is less than the specific judgment threshold, it is judged that the patient does not have amyotrophic lateral sclerosis.

[0100] For example, (1) when the sALS prediction and prognosis assessment model based on multiple proteins pre-set in the computer program is:

[0101] Log(P)=0.002*HBB+0.762*FKBP1A+0.183*CD84+0.589*CAMP+0.172*TLN1+

[0102] 0.744*ZYX+0.047*HBA1+0.209*TPT1-8.476

[0103] , then only the following eight types of data need to be entered: the protein concentration encoded by the HBB gene, the protein concentration encoded by the FKBP1A gene, the protein concentration encoded by the CD84 gene, the protein concentration encoded by the CAMP gene, the protein concentration encoded by the TLN1 gene, the protein concentration encoded by the ZYX gene, the protein concentration encoded by the HBA1 gene, and the protein concentration encoded by the TPT1 gene.

[0104] The judgment threshold of the sALS prediction and prognosis evaluation model based on multiple proteins is 0.528. When the score result is greater than or equal to 0.528, it is judged that the patient has amyotrophic lateral sclerosis, and when the score result is less than 0.528, it is judged that the patient does not have amyotrophic lateral sclerosis.

[0105] To further test the clinical application effect of the model, the model was applied to calculate Log(P) in a new cohort (25 sALS patients and 25 normal controls). In the sALS group, 24 patients with Log(P)>0.528 were diagnosed with ALS, and 1 patient with Log(P)<0.528 was not diagnosed with ALS. In the normal control group, 21 patients with Log(P)<0.528 were not diagnosed with ALS, and 4 patients with Log(P)>0.528 were diagnosed with ALS. The diagnostic sensitivity of the model was 96%, the specificity was 84%, the positive predictive value was 85.71%, and the negative predictive value was 95.45%, which showed good efficacy in distinguishing the disease group from the control group (Fisher's exact test P<0.0001).

[0106] For example, (2) when the sALS prediction and prognosis assessment model based on multiple proteins pre-set in the computer program is:

[0107] Log(P)=0.238*TLN1+0.175*TPT1+1.446*ZYX-5.644

[0108] When , only the following three data need to be input: the protein concentration encoded by the TLN1 gene, the protein concentration encoded by the TPT1 gene, and the protein concentration encoded by the ZYX gene.

[0109] The judgment threshold of the sALS prediction and prognosis evaluation model based on multiple proteins is 0.351. When the score result is greater than or equal to 0.351, it is judged that the patient has amyotrophic lateral sclerosis, and when the score result is less than 0.351, it is judged that the patient does not have amyotrophic lateral sclerosis.

[0110] The output module outputs the prediction and evaluation result according to the judgment result of the prediction and evaluation module: "suffering from amyotrophic lateral sclerosis" or "not suffering from amyotrophic lateral sclerosis".

[0111] It should be noted that the embodiments of the present invention have better practicability and do not impose any form of limitation on the present invention. Any technician familiar with the field may use the technical content disclosed above to change or modify it into an equivalent effective embodiment. However, any modification or equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A method for constructing a prediction and prognosis assessment model for amyotrophic lateral sclerosis, characterized in that: include: Step S1: Based on the test set consisting of ALS patient data and healthy control group data, several potential ALS serum biomarkers were preliminarily screened by label-free quantitative proteomics; Step S2: Based on the validation set consisting of ALS patient data and healthy control group data, the expression levels of the potential ALS serum biomarkers are detected by ELISA detection kit, and then the ALS serum biomarkers with significantly upregulated expression levels are sequentially subjected to principal component dimensionality reduction analysis and interaction analysis, so as to finally screen out several ALS serum biomarkers, and perform receiver operating characteristic analysis to evaluate the sensitivity and specificity of each biomarker in distinguishing sALS patients from healthy controls; Step S3: Based on two or more of the several ALS serum biomarkers, a logistic regression model is constructed by using the likelihood ratio method to generate a sALS prediction and prognosis evaluation model based on multiple proteins, wherein the sALS prediction and prognosis evaluation model based on multiple proteins is used to comprehensively calculate a scoring result according to the concentrations of multiple proteins. If the scoring result is greater than or equal to a specific judgment threshold, it is judged that the patient suffers from amyotrophic lateral sclerosis.

2. The method for constructing an amyotrophic lateral sclerosis prediction and prognosis assessment model according to claim 1, characterized in that: In step S1, false discovery rate analysis was used to identify differentially expressed proteins; principal multivariate component analysis, gene ontology enrichment analysis, pathway enrichment analysis, and protein interaction network analysis were performed using the STRING database to determine the functional clustering of differentially expressed proteins; and several potential ALS serum biomarkers were further screened out based on the degree of expression difference, molecular function, and organ expression specificity.

3. The method for constructing an amyotrophic lateral sclerosis prediction and prognosis assessment model according to claim 2, characterized in that: The candidate ALS serum biomarkers include proteins encoded by FKBP1A, CD84, CAMP, ZYX, HBA1, HBB, TLN1, and TPT1 genes, respectively.

4. The method for constructing an amyotrophic lateral sclerosis prediction and prognosis assessment model according to claim 3, characterized in that: The sALS prediction and prognosis assessment model based on multiple proteins is: Log(P)=0.002*HBB+0.762*FKBP1A+0.183*CD84+0.589*CAMP+0.172*TLN1+ 0.744*ZYX+0.047*HBA1+0.209*TPT1-8.476 Wherein, HBB in the formula represents the protein concentration encoded by the HBB gene, FKBP1A represents the protein concentration encoded by the FKBP1A gene, CD84 represents the protein concentration encoded by the CD84 gene, CAMP represents the protein concentration encoded by the CAMP gene, TLN1 represents the protein concentration encoded by the TLN1 gene, ZYX represents the protein concentration encoded by the ZYX gene, HBA1 represents the protein concentration encoded by the HBA1 gene, and TPT1 represents the protein concentration encoded by the TPT1 gene; The judgment threshold of the sALS prediction and prognosis evaluation model based on multiple proteins is 0.

528.

5. The method for constructing an amyotrophic lateral sclerosis prediction and prognosis assessment model according to claim 3, characterized in that: The sALS prediction and prognosis assessment model based on multiple proteins is: Log(P)=0.238*TLN1+0.175*TPT1+1.446*ZYX-5.644 Wherein, TLN1 in the formula represents the concentration of the protein encoded by the TLN1 gene, TPT1 represents the concentration of the protein encoded by the TPT1 gene, and ZYX represents the concentration of the protein encoded by the ZYX gene; The judgment threshold corresponding to the sALS prediction and prognosis evaluation model based on multiple proteins is 0.351; The model is used for the prediction of early-stage amyotrophic lateral sclerosis.

6. A system for predicting and evaluating the prognosis of amyotrophic lateral sclerosis, characterized in that: include: An input module, the input module is used to obtain the content of multiple candidate ALS serum biomarkers; A prediction and evaluation module, wherein the prediction and evaluation module is used to calculate a score based on a sALS prediction and prognosis evaluation model based on multiple proteins, and when the score result is greater than or equal to a specific judgment threshold, it is judged that the patient suffers from amyotrophic lateral sclerosis, and when the score result is less than the specific judgment threshold, it is judged that the patient does not suffer from amyotrophic lateral sclerosis; An output module is used to output the prediction evaluation result.

7. The amyotrophic lateral sclerosis prediction and prognosis assessment system according to claim 6, characterized in that: The several candidate ALS serum biomarkers are selected from two or more of the proteins encoded by FKBP1A, CD84, CAMP, ZYX, HBA1, HBB, TLN1, and TPT1 genes.

8. The amyotrophic lateral sclerosis prediction and prognosis assessment system according to claim 7, characterized in that: The sALS prediction and prognosis assessment model based on multiple proteins is: Log(P)=0.002*HBB+0.762*FKBP1A+0.183*CD84+0.589*CAMP+0.172*TLN1+ 0.744*ZYX+0.047*HBA1+0.209*TPT1-8.476 Wherein, HBB in the formula represents the protein concentration encoded by the HBB gene, FKBP1A represents the protein concentration encoded by the FKBP1A gene, CD84 represents the protein concentration encoded by the CD84 gene, CAMP represents the protein concentration encoded by the CAMP gene, TLN1 represents the protein concentration encoded by the TLN1 gene, ZYX represents the protein concentration encoded by the ZYX gene, HBA1 represents the protein concentration encoded by the HBA1 gene, and TPT1 represents the protein concentration encoded by the TPT1 gene; The judgment threshold of the sALS prediction and prognosis evaluation model based on multiple proteins is 0.

528.

9. The amyotrophic lateral sclerosis prediction and prognosis assessment system according to claim 7, characterized in that: The sALS prediction and prognosis assessment model based on multiple proteins is: Log(P)=0.238*TLN1+0.175*TPT1+1.446*ZYX-5.644 Wherein, TLN1 in the formula represents the concentration of the protein encoded by the TLN1 gene, TPT1 represents the concentration of the protein encoded by the TPT1 gene, and ZYX represents the concentration of the protein encoded by the ZYX gene; The judgment threshold corresponding to the sALS prediction and prognostic evaluation model is 0.351; The model is used for the prediction of early-stage amyotrophic lateral sclerosis.

Images