Using a machine learning approach to identify key prognostic molecules for esophageal squamous cell carcinoma.

{"sci_version": "1.2", "schema": "https://corpus.bntrg.co.uk/sci/1.0/sci.schema.json", "id": "pmid:34372798", "kind": "science-record", "revision": 8, "content_hash": "sha256:5a5778ef1deb8e35c69616980ef5408e3952e3ca225a168b3836ecc5593fa571", "created_at": "2026-11-01T15:41:47Z", "updated_at": "2026-11-01T22:47:18Z", "generator": {"name": "corpus-assay-engine", "run_id": "reading-sci-2026-11-01"}, "paper": {"state": "complete", "title": "Using a machine learning approach to identify key prognostic molecules for esophageal squamous cell carcinoma.", "doi": "10.1186/s12885-021-08647-1", "pmid": "34372798", "authors": ["Meng-Xiang Li", "Xiao-Meng Sun", "Wei-Gang Cheng", "Hao-Jie Ruan", "Ke Liu", "Pan Chen", "Hai-Jun Xu", "She-Gan Gao", "Xiao-Shan Feng", "Yi-Jun Qi"], "journal": "BMC cancer", "year": 2021, "abstract": "BACKGROUND: A plethora of prognostic biomarkers for esophageal squamous cell carcinoma (ESCC) that have hitherto been reported are challenged with low reproducibility due to high molecular heterogeneity of ESCC. The purpose of this study was to identify the optimal biomarkers for ESCC using machine learning algorithms.\nMETHODS: Biomarkers related to clinical survival, recurrence or therapeutic response of patients with ESCC were determined through literature database searching. Forty-eight biomarkers linked to recurrence or prognosis of ESCC were used to construct a molecular interaction network based on NetBox and then to identify the functional modules. Publicably available mRNA transcriptome data of ESCC downloaded from Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) datasets included GSE53625 and TCGA-ESCC. Five machine learning algorithms, including logical regression (LR), support vector machine (SVM), artificial neural network (ANN), random forest (RF) and XGBoost, were used to develop classifiers for prognostic classification for feature selection. The area under ROC curve (AUC) was used to evaluate the performance of the prognostic classifiers. The importances of identified molecules were ranked by their occurrence frequencies in the prognostic classifiers. Kaplan-Meier survival analysis and log-rank test were performed to determine the statistical significance of overall survival.\nRESULTS: A total of 48 clinically proven molecules associated with ESCC progression were used to construct a molecular interaction network with 3 functional modules comprising 17 component molecules. The 131,071 prognostic classifiers using these 17 molecules were built for each machine learning algorithm. Using the occurrence frequencies in the prognostic classifiers with AUCs greater than the mean value of all 131,071 AUCs to rank importances of these 17 molecules, stratifin encoded by SFN was identified as the optimal prognostic biomarker for ESCC, whose performance was further validated in another 2 independent cohorts.\nCONCLUSION: The occurrence frequencies across various feature selection approaches reflect the degree of clinical importance and stratifin is an optimal prognostic biomarker for ESCC.", "claims": [{"id": "claim-method", "statement": "Kaplan-Meier survival analysis and log-rank test were performed to determine the statistical significance of overall survival.", "kind": "methodological", "target": null, "evidence": [{"asset": "source:pmid:34372798", "locator": "sections[0] \"Abstract\" paragraph 2", "quote": "Kaplan-Meier survival analysis and log-rank test were performed to determine the statistical significance of overall survival."}], "state": "complete"}, {"id": "claim-data-access", "statement": "The data sets of GSE53625 from GEO (https://www.ncbi.nlm.nih.gov/geo/) and 37 ESCC cases with Asian ancestry from TCGA (UCSC Xena, https://xena.ucsc.edu/) are public open and available.", "kind": "methodological", "target": null, "evidence": [{"asset": "source:pmid:34372798", "locator": "sections[24] \"Availability of data and materials\" paragraph 2", "quote": "The data sets of GSE53625 from GEO (https://www.ncbi.nlm.nih.gov/geo/) and 37 ESCC cases with Asian ancestry from TCGA (UCSC Xena, https://xena.ucsc.edu/) are public open and available."}], "state": "complete"}], "gaps": []}, "experiment": {"state": "complete", "objective": null, "steps": [{"id": "read_published_estimate", "order": 1, "name": "Read the published estimate from the source", "kind": "other", "state": "complete", "blocked_on": null, "description": "The claim, method and evidence span published with the source article.", "duration_hours": null, "parameters": {}, "inputs": ["source:pmid:34372798"], "outputs": ["the published estimate and its evidence"], "instruments": [], "reagents": []}, {"id": "obtain_source_data", "order": 2, "name": "Obtain the source data from the named repository", "kind": "data-preparation", "state": "complete", "blocked_on": null, "description": "The GEO series accession resolves and its sample accessions are listed. This lane did not download the dataset or read its licence.", "duration_hours": null, "parameters": {}, "inputs": ["data:geo:GSE53625"], "outputs": ["the source data with its repository provenance"], "instruments": [], "reagents": []}, {"id": "run_analysis", "order": 3, "name": "Run the published analysis", "kind": "analysis", "state": "complete", "blocked_on": null, "description": "Apply the paper's statistical method to the resolved series and compare the result with the published target.", "duration_hours": null, "parameters": {}, "inputs": ["data:geo:GSE53625"], "outputs": ["the replicated result and its comparison"], "instruments": [], "reagents": []}], "materials": [], "instruments": [], "reagents": [], "conditions": {}, "gaps": []}, "data": {"state": "partial", "assets": [{"id": "source:pmid:34372798", "role": "manuscript", "format": "xml", "media_type": "application/xml", "uri": "https://doi.org/10.1186/s12885-021-08647-1", "sha256": "sha256:a71008efbf5729c6b73d3520fffad9111ec13b952e86229413e8de407ed892e1", "bytes": 41500, "access": "open", "state": "resolved", "reader": null, "reader_status": "reader_required", "identifier": "10.1186/s12885-021-08647-1", "version": null, "note": "Full article fetched from PMC; the checksum covers the rendered section body, not the original XML bytes.", "structure": {"pmcid": "PMC8351329", "source": "europe_pmc_fulltext", "coverage": "whole_body", "text_hash_scope": "rendered_section_body"}}, {"id": "data:geo:GSE53625", "role": "raw", "format": "geo-series", "media_type": "application/vnd.ncbi.geo-series+json", "uri": "https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53625", "sha256": null, "bytes": null, "access": "open", "state": "resolved", "reader": null, "reader_status": "reader_required", "identifier": "GSE53625", "version": null, "note": "The data sets of GSE53625 from GEO (https://www.ncbi.nlm.nih.gov/geo/) and 37 ESCC cases with Asian ancestry from TCGA (UCSC Xena, https://xena.ucsc.edu/) are public open and available.", "structure": {"repository": "GEO", "access_route": "GEO accession URL", "access_statement_verbatim": "The data sets of GSE53625 from GEO (https://www.ncbi.nlm.nih.gov/geo/) and 37 ESCC cases with Asian ancestry from TCGA (UCSC Xena, https://xena.ucsc.edu/) are public open and available.", "access_statement_kind": "data_availability_statement", "sample_accessions": ["GSM1296956", "GSM1296957", "GSM1296958", "GSM1296959", "GSM1296960", "GSM1296961", "GSM1296962", "GSM1296963", "GSM1296964", "GSM1296965", "GSM1296966", "GSM1296967", "GSM1296968", "GSM1296969", "GSM1296970", "GSM1296971", "GSM1296972", "GSM1296973", "GSM1296974", "GSM1296975", "GSM1296976", "GSM1296977", "GSM1296978", "GSM1296979", "GSM1296980", "GSM1296981", "GSM1296982", "GSM1296983", "GSM1296984", "GSM1296985", "GSM1296986", "GSM1296987", "GSM1296988", "GSM1296989", "GSM1296990", "GSM1296991", "GSM1296992", "GSM1296993", "GSM1296994", "GSM1296995", "GSM1296996", "GSM1296997", "GSM1296998", "GSM1296999", "GSM1297000", "GSM1297001", "GSM1297002", "GSM1297003", "GSM1297004", "GSM1297005", "GSM1297006", "GSM1297007", "GSM1297008", "GSM1297009", "GSM1297010", "GSM1297011", "GSM1297012", "GSM1297013", "GSM1297014", "GSM1297015", "GSM1297016", "GSM1297017", "GSM1297018", "GSM1297019", "GSM1297020", "GSM1297021", "GSM1297022", "GSM1297023", "GSM1297024", "GSM1297025", "GSM1297026", "GSM1297027", "GSM1297028", "GSM1297029", "GSM1297030", "GSM1297031", "GSM1297032", "GSM1297033", "GSM1297034", "GSM1297035", "GSM1297036", "GSM1297037", "GSM1297038", "GSM1297039", "GSM1297040", "GSM1297041", "GSM1297042", "GSM1297043", "GSM1297044", "GSM1297045", "GSM1297046", "GSM1297047", "GSM1297048", "GSM1297049", "GSM1297050", "GSM1297051", "GSM1297052", "GSM1297053", "GSM1297054", "GSM1297055", "GSM1297056", "GSM1297057", "GSM1297058", "GSM1297059", "GSM1297060", "GSM1297061", "GSM1297062", "GSM1297063", "GSM1297064", "GSM1297065", "GSM1297066", "GSM1297067", "GSM1297068", "GSM1297069", "GSM1297070", "GSM1297071", "GSM1297072", "GSM1297073", "GSM1297074", "GSM1297075", "GSM1297076", "GSM1297077", "GSM1297078", "GSM1297079", "GSM1297080", "GSM1297081", "GSM1297082", "GSM1297083", "GSM1297084", "GSM1297085", "GSM1297086", "GSM1297087", "GSM1297088", "GSM1297089", "GSM1297090", "GSM1297091", "GSM1297092", "GSM1297093", "GSM1297094", "GSM1297095", "GSM1297096", "GSM1297097", "GSM1297098", "GSM1297099", "GSM1297100", "GSM1297101", "GSM1297102", "GSM1297103", "GSM1297104", "GSM1297105", "GSM1297106", "GSM1297107", "GSM1297108", "GSM1297109", "GSM1297110", "GSM1297111", "GSM1297112", "GSM1297113", "GSM1297114", "GSM1297115", "GSM1297116", "GSM1297117", "GSM1297118", "GSM1297119", "GSM1297120", "GSM1297121", "GSM1297122", "GSM1297123", "GSM1297124", "GSM1297125", "GSM1297126", "GSM1297127", "GSM1297128", "GSM1297129", "GSM1297130", "GSM1297131", "GSM1297132", "GSM1297133", "GSM1297134", "GSM1297135", "GSM1297136", "GSM1297137", "GSM1297138", "GSM1297139", "GSM1297140", "GSM1297141", "GSM1297142", "GSM1297143", "GSM1297144", "GSM1297145", "GSM1297146", "GSM1297147", "GSM1297148", "GSM1297149", "GSM1297150", "GSM1297151", "GSM1297152", "GSM1297153", "GSM1297154", "GSM1297155", "GSM1297156", "GSM1297157", "GSM1297158", "GSM1297159", "GSM1297160", "GSM1297161", "GSM1297162", "GSM1297163", "GSM1297164", "GSM1297165", "GSM1297166", "GSM1297167", "GSM1297168", "GSM1297169", "GSM1297170", "GSM1297171", "GSM1297172", "GSM1297173", "GSM1297174", "GSM1297175", "GSM1297176", "GSM1297177", "GSM1297178", "GSM1297179", "GSM1297180", "GSM1297181", "GSM1297182", "GSM1297183", "GSM1297184", "GSM1297185", "GSM1297186", "GSM1297187", "GSM1297188", "GSM1297189", "GSM1297190", "GSM1297191", "GSM1297192", "GSM1297193", "GSM1297194", "GSM1297195", "GSM1297196", "GSM1297197", "GSM1297198", "GSM1297199", "GSM1297200", "GSM1297201", "GSM1297202", "GSM1297203", "GSM1297204", "GSM1297205", "GSM1297206", "GSM1297207", "GSM1297208", "GSM1297209", "GSM1297210", "GSM1297211", "GSM1297212", "GSM1297213", "GSM1297214", "GSM1297215", "GSM1297216", "GSM1297217", "GSM1297218", "GSM1297219", "GSM1297220", "GSM1297221", "GSM1297222", "GSM1297223", "GSM1297224", "GSM1297225", "GSM1297226", "GSM1297227", "GSM1297228", "GSM1297229", "GSM1297230", "GSM1297231", "GSM1297232", "GSM1297233", "GSM1297234", "GSM1297235", "GSM1297236", "GSM1297237", "GSM1297238", "GSM1297239", "GSM1297240", "GSM1297241", "GSM1297242", "GSM1297243", "GSM1297244", "GSM1297245", "GSM1297246", "GSM1297247", "GSM1297248", "GSM1297249", "GSM1297250", "GSM1297251", "GSM1297252", "GSM1297253", "GSM1297254", "GSM1297255", "GSM1297256", "GSM1297257", "GSM1297258", "GSM1297259", "GSM1297260", "GSM1297261", "GSM1297262", "GSM1297263", "GSM1297264", "GSM1297265", "GSM1297266", "GSM1297267", "GSM1297268", "GSM1297269", "GSM1297270", "GSM1297271", "GSM1297272", "GSM1297273", "GSM1297274", "GSM1297275", "GSM1297276", "GSM1297277", "GSM1297278", "GSM1297279", "GSM1297280", "GSM1297281", "GSM1297282", "GSM1297283", "GSM1297284", "GSM1297285", "GSM1297286", "GSM1297287", "GSM1297288", "GSM1297289", "GSM1297290", "GSM1297291", "GSM1297292", "GSM1297293", "GSM1297294", "GSM1297295", "GSM1297296", "GSM1297297", "GSM1297298", "GSM1297299", "GSM1297300", "GSM1297301", "GSM1297302", "GSM1297303", "GSM1297304", "GSM1297305", "GSM1297306", "GSM1297307", "GSM1297308", "GSM1297309", "GSM1297310", "GSM1297311", "GSM1297312", "GSM1297313"], "sample_count": 358, "group_sizes": null, "group_sizes_basis": null, "platform": "18109", "taxon": "Homo sapiens", "title": "LncRNA profile study reveals a three-lncRNA signature associated with the survival of esophageal squamous cell carcinoma patients", "resolution_checked_at": "2026-11-01T15:39:54Z", "resolution_check": "identifier_resolution_only_no_download_no_licence_read", "downloaded": false, "licence_read": false}}, {"id": "asset:table:pmid:34372798:1", "role": "table", "format": "jats-table", "media_type": "application/vnd.jats.table+json", "uri": null, "sha256": "sha256:a2620f7c578ad22dd99ed5cd6398f39dc51ffab72cc02da86cba6883cab25c84", "bytes": 8414, "access": "open", "state": "verified", "reader": "read_table", "reader_status": "available", "identifier": "section[13]/table-wrap[1]#Tab1", "version": null, "note": "Thirty-eight studies reporting 48 molecules that were associated with clinical survival, recurrence or therapeutic outcome of ESCC patients", "structure": {"pmcid": "PMC8351329", "section_index": 13, "section_title": "Prognostic biomarkers of esophageal squamous cell carcinoma", "row_count": 39, "column_count": 5}, "table": {"label": "Table 1", "caption": "Thirty-eight studies reporting 48 molecules that were associated with clinical survival, recurrence or therapeutic outcome of ESCC patients", "rows": [["Biomarkers", "Prognosis/recurrence/therapeutic outcome", "PMID", "Sample sizes", "Conclusions"], ["BRCA1", "Prognosis and chemoradiotherapy", "23,326,344", "144", "Low BRCA1 expression was an independent prognostic factor in cisplatin-based chemotherapy (HR 0.29, 95% CI 0.12–0.71; P = 0.007) or chemoradiotherapy (HR 0.12, 95% CI 0.04–0.37; P < 0.001) group."], ["CCNA2", "Chemotherapy", "23,205,070", "48", "The expression of cyclin A was an independent prognosis factor in patients with ESCC following paclitaxel-based chemotherapy."], ["CCND1", "Prognosis and radiochemotherapy", "9,988,238", "172", "Patients with cyclin D1-positive carcinomas showed significantly worse overall survival than patients with cyclin D1-negative carcinomas (HR 2.14, 95% CI 1.134–3.42; P = 0.0038)"], ["CD163 & CD68", "Prognosis and neoadjuvant chemotherapy", "25,752,960", "210", "High infiltration of CD68+ macrophages and CD163+ macrophages was significantly associated with poor prognosis for patients undergoing neoadjuvant chemotherapy (P = 0.057, P = 0.003)."], ["CD274", "Prognosis and chemoradiotherapy", "26,623,522", "45", "The higher PD-L1 H-scores had poorer overall survival (median 16.7 versus 32.9 months, P = 0.02) than those with lower H-scores. (HR 2.29, 95% CI 1.12–4.69; P = 0.023)"], ["CD44 & PROM1", "Prognosis", "27,748,881", "47", "Patients with strong expression of CD44 or CD133 and those with a high ratio of CD133-positive tumor cells showed significantly poor prognosis regardless of the effect of chemotherapy. PROM1 (HR 5.05, 95% CI 1.12–4.69; P = 0.023)"], ["CDKN2A", "neoadjuvant chemotherapy", "26,514,506", "101", "ESCC tumors that were found positive for p16 expression appeared to fall into responders group rather than non responders (P = 0.008) and reported with less mortality (P = 0.048)."], ["CEA & KRT19", "Chemoradiotherapy", "19,863,186", "84", "CEA may be helpful in predicting the responsiveness in ESCC of primary lesions to CRT, with the effective rates (CR + PR) in CEA high and low groups of 58.3% (14/24) and 93.3% (56/60), respectively (P = 0.013 and 0.013)."], ["EGFR", "Chemoradiotherapy", "17,940,077", "62", "The difference in the CR rate between EGFR positive and -negative groups was significant (CR rate: 62% vs. 34%; P = 0.037)."], ["CRP", "Chemoradiotherapy", "21,224,533", "36", "Serum CRP can predict of CRT response with an accuracy of 75%."], ["ERCC1", "Chemotherapy", "23,263,828", "46", "Patients with ERCC1 negative tumors had a higher treatment response than the ERCC1 positive group (radiological response rates; 92.3% vs.50%, P = 0.013)."], ["FAM84B", "Neoadjuvant chemoradiation", "25,980,316", "21", "The fold-change of circulating FAM84B mRNA expression can predict the pCR with an AUC of 0.73."], ["FDXR", "Prognosis and Neoadjuvant chemoradiation", "26,637,858", "50", "Fdxr was significantly correlated with postoperative outcomes and an independent prognostic factor (HR 4.950, 95% CI 1.603–15.38; 0.012)."], ["HOXC6 & HOXC8", "Prognosis", "24,525,058", "274", "HOXC6 and HOXC8 were independent prognostic factors in patients with ESCC. HOXC6: (HR 1.341, 95% CI 0.895–2.010; P = 0.045); HOXC8: (HR 1.657, 95% CI 1.146–2.395; P = 0.007)."], ["IL6R", "Prognosis", "23,648,090", "218", "The sIL6R level was one of several significant independent predictors of an unfavorable outcome. (HR 3.20, 95% CI 1.34–7.53; P = 0.008)"], ["MDM2 & MKI67", "Prognosis and chemoradiotherapy", "25,880,782", "79", "MDM2 and p16 are predictive markers for chemoradioresistance in cStageIII ESCC and Ki-67 is a prognostic marker following dCRT in cStageIII ESCC."], ["MLH1", "Prognosis", "18,053,639", "51", "The expression of hMLH1 is a potential marker of tumor response and survival."], ["MMS19", "Chemoradiotherapy", "25,892,874", "103", "High cytoplasmic MMS19 expression was associated with a good response to chemoradiotherapy (OR 11.5, 95% CI: 3.0–44.5; P < 0.001)."], ["MT3", "Prognosis", "16,351,731", "64", "Esophageal squamous cell carcinomas with negative p53, positive CDC25B, and negative MT expressions respond well to CRT."], ["MUC13 & MUC20", "Prognosis and neoadjuvant chemotherapy", "26,323,930", "186", "The median survival time of patients with low MUC13/high MUC20 expression was significantly shorter than that of patients with high MUC13/low MUC20 expression (27.7 months vs. 59.5 months, P = 0.021; HR 0.531, 95% CI: 0.299–0.944; P = 0.031)."], ["MUC4", "neoadjuvant chemotherapy", "26,673,820", "186", "Low expression of MUC4 and MUC20 in resection samples was significantly correlated with better TRG (tumor regression grade). MUC4 and MUC20 were identified as potential biomarkers for predicting the efficacy of neoadjuvant chemotherapy in ESCC patients."], ["NOTCH1 & PIK3CA", "Prognosis and Chemotherapy", "26,528,858", "104", "NOTCH1 mutations was correlated with shorter survival times and failed to respond to chemotherapy, whereas PIK3CA mutations pointed to better responses to chemotherapy and longer survival times than patients without PIK3CA mutations."], ["PTGS2", "Prognosis and Chemoradiotherapy", "21,437,756", "58", "Negative or weak expression of PTGS2 was correlated significantly with CRT response (OR 6.296, 95% CI 1.58–25.096; P = 0.010)."], ["PTPN6", "Prognosis", "32,536,826", "184", "Elevated PTPN6 expression indicated longer OS (HR 1.123, 95% CI: 0.565–2.230; P = 0.741)."], ["RAD51", "Prognosis and recurrence", "24,065,387", "89", "Rad51 expression in ESCC was associated with poor survival (P = 0.0324) and recurrence (P = 0.0171)."], ["REG1A", "Prognosis", "23,645,481", "177", "REG1A expression was a significant prognostic factor (HR 3.095, 95% CI: 1.569–5.943; P = 0.0015)."], ["SFN", "Chemoradiation therapy", "15,999,354", "62", "SFN-positive expressions were closely related to the response to CRT."], ["Prognosis", "24,743,601", "278", "Downregulation of 14–3-3σpredicts poor survival, suggesting that 14–3-3σmay be a biomarker for early detection of high-risk subjects and diagnosis of ESCC. (HR 0.466, 95% CI 0.251–0.866; P = 0.016)."], ["Prognosis", "20,108,042", "148", "Reduced stratifin expression, T4 stage, lymph node metastasis, and distant metastasis were independent risk factors for worse prognosis in ESCC patients."], ["SGTA", "Prognosis", "23,939,810", "120", "SGTA expression indicated poor prognosis (RR 3.513, 95% CI: 2.161–9.791; P = 0.016)."], ["TGFB1 & VEGFA", "Prognosis", "24,623,035", "79", "VEGFA and TGFB1 were significantly associated with pathological response and/or DFS, and may be used to predict pathological response and survivals for ESCC patients receiving combined modality therapy."], ["TP53 & RRM2B", "Prognosis and chemoradiotherapy", "15,655,547", "62", "p53 or p53R2 (RRM2B) expression was correlated with a favorable response to CRT (P = 0.0001 or 0.041 clinical, P = 0.016 or 0.0018 histological, respectively; TP53, RR 2.688, 95% CI: 1.157–6.250; P = 0.0011. RRM2B, RR 2.469, 95% CI: 1.164–5.235; P = 0.0057)."], ["Prognosis", "25,135,238", "36", "The median tumor associated survival was 34.2 months for patients with normal TP53, compared with 8.9 months for those with mutant TP53. The latter had a 3-fold higher risk of death (HR 3.01, 95% CI 1.359–6.86; P = 0.005)."], ["Prognosis", "10,414,702", "42", "The current study indicated that p53 mutation of tumor tissues might be a prognostic factor for esophageal squamous cell carcinoma cases and one of the risk factors for its recurrence."], ["Chemotherapy", "19,941,080", "97", "Patients with mutations in p53 therefore showed significantly poorer prognosis than those without mutant p53."], ["RAC3 & TRAM1", "Prognosis and chemoradiotherapy", "19,552,757", "98", "Overexpression of AIB1/RAC3/ TRAM1 is a useful predictor of CRT resistance and an independent molecular marker of poor prognosis for ESCC patients."], ["ALDH1A1, ALDH1A2, ALDH1A3, ALDH1B1, ALDH1L1, ALDH1L2", "Prognosis and recurrence", "22,847,125", "152", "ALDH1 was a predictor of postoperative recurrence and prognosis in ESCC, and CD44 might be a predictor of recurrence and prognosis."], ["PITX2", "Prognosis and chemoradiotherapy", "23,132,660", "454", "High expression of PITX2 was associated with poor disease-specific survival (HR 1.732, 95% CI 1.133–2.646; P = 0.011) in ESCC."]], "row_count": 39, "column_count": 5, "numeric_cells": 117, "header_rows": 1, "locator": "section[13]/table-wrap[1]#Tab1"}}, {"id": "asset:table:pmid:34372798:2", "role": "table", "format": "jats-table", "media_type": "application/vnd.jats.table+json", "uri": null, "sha256": "sha256:dcbaf7bb4d3876419e4ac99adfd6e21caceb3ca5cf42354a1a2ed2ae3aaddbbc", "bytes": 528, "access": "open", "state": "verified", "reader": "read_table", "reader_status": "available", "identifier": "section[15]/table-wrap[1]#Tab2", "version": null, "note": "The top 5 important molecules identified by each machine learning algorithm", "structure": {"pmcid": "PMC8351329", "section_index": 15, "section_title": "Prognostic classification using 5 machine learning algorithms", "row_count": 7, "column_count": 6}, "table": {"label": "Table 2", "caption": "The top 5 important molecules identified by each machine learning algorithm", "rows": [["Moleculerank", "Weak classifiers", "Strong classifiers"], ["LR", "SVM", "ANN", "RF", "XGBoost"], ["1", "CD44", "CD44", "SFN", "SFN", "SFN"], ["2", "RAC3", "TRAM1", "CCND1", "PTGS2", "PIK3CA"], ["3", "TP53", "SFN", "CD44", "MDM2", "CD44"], ["4", "EGFR", "PTGS2", "PTGS2", "PIK3CA", "VEGFA"], ["5", "SFN", "VEGFA", "MDM2", "RAC3", "PTGS2"]], "row_count": 7, "column_count": 6, "numeric_cells": 10, "header_rows": 2, "locator": "section[15]/table-wrap[1]#Tab2"}}], "gaps": ["asset:data:geo:GSE53625 not downloaded", "asset:data:geo:GSE53625 licence not read"]}, "analysis": {"state": "partial", "methods": [{"id": "method-1", "family": "survival_analysis", "specific": "kaplan_meier", "label": "Survival analysis", "dsl_operations": ["survival_km"], "missing_operations": [], "expressible": true, "parameters": {}, "data_shapes": ["time_to_event_with_censoring"], "evidence": {"score": 1.0, "band": "verbatim", "quote": "Kaplan-Meier survival analysis and log-rank test were performed to determine the statistical significance of overall survival."}}, {"id": "method-2", "family": "group_comparison", "specific": "student_t_test", "label": "Group comparison", "dsl_operations": [], "missing_operations": ["student_t"], "expressible": false, "parameters": {}, "data_shapes": ["two_or_more_groups_continuous"], "evidence": {"score": 0.8, "band": "high", "quote": "Student t-test"}}, {"id": "method-3", "family": "survival_analysis", "specific": "log_rank_test", "label": "Survival analysis", "dsl_operations": ["logrank_test"], "missing_operations": [], "expressible": true, "parameters": {}, "data_shapes": ["time_to_event_with_censoring"], "evidence": {"score": 0.7, "band": "medium", "quote": "log-rank"}}, {"id": "method-4", "family": "machine_learning", "specific": "random_forest", "label": "Machine learning", "dsl_operations": [], "missing_operations": ["machine_learning_random_forest"], "expressible": false, "parameters": {}, "data_shapes": ["labelled_feature_matrix"], "evidence": {"score": 0.7, "band": "medium", "quote": "random forest"}}, {"id": "method-5", "family": "machine_learning", "specific": "support_vector_machine", "label": "Machine learning", "dsl_operations": ["train_classifier", "cross_validate", "roc_auc"], "missing_operations": [], "expressible": true, "parameters": {}, "data_shapes": ["labelled_feature_matrix"], "evidence": {"score": 0.7, "band": "medium", "quote": "support vector machine"}}, {"id": "method-6", "family": "machine_learning", "specific": "gradient_boosting", "label": "Machine learning", "dsl_operations": ["train_classifier", "cross_validate", "roc_auc"], "missing_operations": [], "expressible": true, "parameters": {}, "data_shapes": ["labelled_feature_matrix"], "evidence": {"score": 0.7, "band": "medium", "quote": "XGBoost"}}, {"id": "method-7", "family": "regression_modelling", "specific": "ols", "label": "Regression modelling", "dsl_operations": ["ols_fit"], "missing_operations": [], "expressible": true, "parameters": {}, "data_shapes": ["predictors_and_outcome"], "evidence": {"score": 0.8, "band": "high", "quote": "linear regression"}}, {"id": "method-8", "family": "diagnostic_evaluation", "specific": "roc", "label": "Diagnostic test evaluation", "dsl_operations": ["confusion_counts", "roc_curve", "sensitivity_specificity", "solve_confusion_matrix", "classification_metrics", "published_precision_envelope"], "missing_operations": [], "expressible": true, "parameters": {}, "data_shapes": ["reference_standard_and_test"], "evidence": {"score": 0.7, "band": "medium", "quote": "ROC curve"}}], "targets": [], "gaps": ["method.reader_differs_from_extractor"]}, "capability": {"state": "complete", "requirements": [{"capability_id": "capability:computational_analysis", "label": "Computational analysis platform", "assay_families": ["meta_analysis", "machine_learning", "survival_analysis", "statistical_regression", "pca", "molecular_docking", "network_analysis", "differential_expression", "gwas"], "availability_state": "owned", "tier_unlocked": "R1", "candidate": "Automated Assays Laravel application and simulated adapters"}], "gaps": []}, "readiness": {"state": "complete", "tier": "R1", "blocking_axis": "none", "blocked_on": "nothing:replicable_now", "axes": {"method": {"tier": "R1", "expressible": true, "required_operations": ["survival_km"], "missing_operations": [], "basis": "The paper names the method it used and the corpus's method extractor maps it onto the DSL catalogue: kaplan_meier."}, "capability": {"tier": "R1", "capability_id": "capability:computational_analysis", "availability_state": "owned", "basis": "The computational analysis is owned. The physical bench the paper used is not, so a physical validation cannot be run here."}, "data": {"tier": "R1", "reachable": true, "accessions": ["GSE53625"], "resolved_accessions": ["GSE53625"], "statements": ["The data sets of GSE53625 from GEO (https://www.ncbi.nlm.nih.gov/geo/) and 37 ESCC cases with Asian ancestry from TCGA (UCSC Xena, https://xena.ucsc.edu/) are public open and available."], "basis": "The paper names GSE53625 in Gene Expression Omnibus (GEO) and the accession resolves in the repository. No dataset was downloaded and no licence was read."}}, "can_run": {"dsl_operations": ["survival_km"], "assets": ["source:pmid:34372798", "data:geo:GSE53625", "asset:table:pmid:34372798:1", "asset:table:pmid:34372798:2"], "steps": ["run_analysis"]}, "must_build": {"dsl_operations": [], "converters": [], "readers": [], "capabilities": [], "data_gaps": ["data:geo:GSE53625"]}, "gaps": []}, "review": {"state": "partial", "machine_critique": {"kind": "extraction_review", "reader": {"model": "MiniMax-M3.1-Flash", "provider": "minimax", "provider_family": "minimax", "read_at": "2026-10-31T18:33:33Z", "independence": "single_reader"}, "evidence_check": {"method": "verbatim", "availability": "verbatim"}, "calls": 3, "usage": {"prompt_tokens": 33410, "completion_tokens": 1456, "cached_tokens": 776}}, "human_critique": null, "stages": [{"id": "extraction", "state": "complete", "reader": "MiniMax-M3.1-Flash", "provider": "minimax"}], "gaps": ["review.independent_reader", "review.machine_critique", "review.human_critique"]}, "execution": {"state": "absent", "status": "not_run", "mode": null, "runs": [], "gaps": ["execution.not_run"]}, "provenance": {"entities": [{"id": "pmid:34372798", "type": "paper", "label": "pmid:34372798"}, {"id": "source:pmid:34372798", "type": "manuscript", "label": "PMID 34372798 full text"}, {"id": "data:geo:GSE53625", "type": "dataset", "label": "GEO GSE53625"}], "activities": [{"id": "activity-reading-sci-2026-11-01", "type": "classification", "used": [{"id": "pmid:34372798", "type": "paper", "label": "pmid:34372798"}, {"id": "source:pmid:34372798", "type": "manuscript", "label": "PMID 34372798 full text"}, {"id": "data:geo:GSE53625", "type": "dataset", "label": "GEO GSE53625"}], "agent": "agent:corpus-assay-engine"}, {"id": "activity-read-34372798", "type": "extraction", "used": ["source:pmid:34372798"], "agent": "agent:minimax-reader"}, {"id": "activity-resolve-GSE53625", "type": "accession_resolution", "used": ["data:geo:GSE53625"], "agent": "agent:corpus-assay-engine"}], "agents": [{"id": "agent:corpus-assay-engine", "type": "software", "label": "Corpus assay engine"}, {"id": "agent:minimax-reader", "type": "software", "label": "MiniMax literature reader"}], "source_run_id": "reading-sci-2026-11-01", "dsl_catalogue_fingerprint": "343fb30756ba6ec3", "capability_table_fingerprint": "2cb3e42f047d2ecf", "snapshot_id": "ba34b49490d9bfaf"}, "lineage": {"parent_id": "pmid:34372798", "parent_hash": "sha256:2a0570ed67a6f9c16bda0b77d9f53380f0b4a06617fb7a5e68ddb181b24cac6c", "fork_of": null, "generation": 7}, "gaps": ["asset:data:geo:GSE53625 licence not read", "asset:data:geo:GSE53625 not downloaded", "execution.not_run", "method.reader_differs_from_extractor", "review.human_critique", "review.independent_reader", "review.machine_critique"], "assays": [{"role": "data_generation", "definition_id": "western_blot", "definition_version": "99e1195c7d6f", "definition_digest": null, "required_operations": [], "support_state": "partial", "source_hashes": {"assay_lexicon": "99e1195c7d6f"}, "confidence": 0.93, "evidence": ["t. The relationship between the abundance of western blot and the expression level of SFN was analyzed"], "order_basis": "paper_text_order:2819", "id": "assay:1"}, {"role": "analysis", "definition_id": "survival_analysis", "definition_version": "f2132e96ae6c7a4e", "definition_digest": null, "required_operations": ["survival_km"], "support_state": "supported", "source_hashes": {"dsl_catalogue": "f2132e96ae6c7a4e"}, "confidence": 1.0, "evidence": ["Kaplan-Meier survival analysis and log-rank test were performed to determine the statistical significance of overall survival."], "order_basis": "consumes_earlier_output", "id": "assay:2"}]}