Identification of gene profiles related to the development of oral cancer using a deep learning technique.

{"sci_version": "1.2", "schema": "https://corpus.bntrg.co.uk/sci/1.0/sci.schema.json", "id": "pmid:36849997", "kind": "science-record", "revision": 3, "content_hash": "sha256:f0c3de2502d51df731cfb52063e6c6610b97f846fc1612557428f3971d270959", "created_at": "2026-11-01T19:01:29Z", "updated_at": "2026-11-01T22:08:57Z", "generator": {"name": "corpus-assay-engine", "run_id": "reading-sci-2026-11-01"}, "paper": {"state": "complete", "title": "Identification of gene profiles related to the development of oral cancer using a deep learning technique.", "doi": "10.1186/s12920-023-01462-6", "pmid": "36849997", "authors": ["Leili Tapak", "Mohammad Kazem Ghasemi", "Saeid Afshar", "Hossein Mahjub", "Alireza Soltanian", "Hassan Khotanlou"], "journal": "BMC medical genomics", "year": 2023, "abstract": "BACKGROUND: Oral cancer (OC) is a debilitating disease that can affect the quality of life of these patients adversely. Oral premalignant lesion patients have a high risk of developing OC. Therefore, identifying robust survival subgroups among them may significantly improve patient therapy and care. This study aimed to identify prognostic biomarkers that predict the time-to-development of OC and survival stratification for patients using state-of-the-art machine learning and deep learning.\nMETHODS: Gene expression profiles (29,096 probes) related to 86 patients from the GSE26549 dataset from the GEO repository were used. An autoencoder deep learning neural network model was used to extract features. We also used a univariate Cox regression model to select significant features obtained from the deep learning method (P < 0.05). High-risk and low-risk groups were then identified using a hierarchical clustering technique based on 100 encoded features (the number of units of the encoding layer, i.e., bottleneck of the network) from autoencoder and selected by Cox proportional hazards model and a supervised random forest (RF) classifier was used to identify gene profiles related to subtypes of OC from the original 29,096 probes.\nRESULTS: Among 100 encoded features extracted by autoencoder, seventy features were significantly related to time-to-OC-development, based on the univariate Cox model, which was used as the inputs for the clustering of patients. Two survival risk groups were identified (P value of log-rank test = 0.003) and were used as the labels for supervised classification. The overall accuracy of the RF classifier was 0.916 over the test set, yielded 21 top genes (FUT8-DDR2-ATM-CD247-ETS1-ZEB2-COL5A2-GMAP7-CDH1-COL11A2-COL3A1-AHR-COL2A1-CHORDC1-PTP4A3-COL1A2-CCR2-PDGFRB-COL1A1-FERMT2-PIK3CB) associated with time to developing OC, selected among the original 29,096 probes.\nCONCLUSIONS: Using deep learning, our study identified prominent transcriptional biomarkers in determining high-risk patients for developing oral cancer, which may be prognostic as significant targets for OC therapy. The identified genes may serve as potential targets for oral cancer chemoprevention. Additional validation of these biomarkers in experimental prospective and retrospective studies will launch them in OC clinics.", "claims": [{"id": "claim-method", "statement": "The extracted features from the autoencoder were examined through the univariate Cox proportional hazards (Cox-PH) model [30] to select significant features (P<0.05).", "kind": "methodological", "target": null, "evidence": [{"asset": "source:pmid:36849997", "locator": "sections[5] \"Hierarchical clustering and feature selection\" paragraph 2", "quote": "The extracted features from the autoencoder were examined through the univariate Cox proportional hazards (Cox-PH) model [30] to select significant features (P<0.05)."}], "state": "complete"}, {"id": "claim-data-access", "statement": "The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request.", "kind": "methodological", "target": null, "evidence": [{"asset": "source:pmid:36849997", "locator": "sections[19] \"Availability of data and materials\" paragraph 2", "quote": "The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request."}], "state": "complete"}, {"id": "claim-target-result", "statement": "The autoencoder-based pipeline identified two survival-risk groups that were significantly different by the log-rank test (P = 0.003).", "kind": "quantitative", "target": {"value": 0.003, "unit": "p-value", "tolerance": null, "source": "Furthermore the survival analysis on the entire data (86 patients) showed that the survival curves in the two identified clusters (Fig. 2) were statistically different (log-rank test P = 0.003)."}, "evidence": [{"asset": "source:pmid:36849997", "locator": "sections[0] \"Abstract\" paragraph 3", "quote": "Two survival risk groups were identified (P value of log-rank test = 0.003) and were used as the labels for supervised classification."}, {"asset": "source:pmid:36849997", "locator": "sections[10] \"Results\" paragraph 3", "quote": "Furthermore the survival analysis on the entire data (86 patients) showed that the survival curves in the two identified clusters (Fig. 2) were statistically different (log-rank test P = 0.003)."}], "state": "complete"}], "gaps": []}, "experiment": {"state": "complete", "objective": "The autoencoder-based pipeline identified two survival-risk groups that were significantly different by the log-rank test (P = 0.003).", "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:36849997"], "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:GSE26549"], "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:GSE26549"], "outputs": ["the replicated result and its comparison"], "instruments": [], "reagents": []}], "materials": [], "instruments": [], "reagents": [], "conditions": {}, "gaps": []}, "data": {"state": "partial", "assets": [{"id": "source:pmid:36849997", "role": "manuscript", "format": "xml", "media_type": "application/xml", "uri": "https://doi.org/10.1186/s12920-023-01462-6", "sha256": "sha256:86c52c586d60e2e08dd3051e4137c2363b1079d38709749b70767517ab215cd5", "bytes": 45280, "access": "open", "state": "resolved", "reader": null, "reader_status": "reader_required", "identifier": "10.1186/s12920-023-01462-6", "version": null, "note": "Full article fetched from PMC; the checksum covers the rendered section body, not the original XML bytes.", "structure": {"pmcid": "PMC9972685", "source": "europe_pmc_fulltext", "coverage": "whole_body", "text_hash_scope": "rendered_section_body"}}, {"id": "data:geo:GSE26549", "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=GSE26549", "sha256": null, "bytes": null, "access": "open", "state": "resolved", "reader": null, "reader_status": "reader_required", "identifier": "GSE26549", "version": null, "note": "The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request.", "structure": {"repository": "GEO", "access_route": "GEO accession URL", "access_statement_verbatim": "The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request.", "access_statement_kind": "data_availability_statement", "sample_accessions": ["GSM652761", "GSM652762", "GSM652763", "GSM652764", "GSM652765", "GSM652766", "GSM652767", "GSM652768", "GSM652769", "GSM652770", "GSM652771", "GSM652772", "GSM652773", "GSM652774", "GSM652775", "GSM652776", "GSM652777", "GSM652778", "GSM652779", "GSM652780", "GSM652781", "GSM652782", "GSM652783", "GSM652784", "GSM652785", "GSM652786", "GSM652787", "GSM652788", "GSM652789", "GSM652790", "GSM652791", "GSM652792", "GSM652793", "GSM652794", "GSM652795", "GSM652796", "GSM652797", "GSM652798", "GSM652799", "GSM652800", "GSM652801", "GSM652802", "GSM652803", "GSM652804", "GSM652805", "GSM652806", "GSM652807", "GSM652808", "GSM652809", "GSM652810", "GSM652811", "GSM652812", "GSM652813", "GSM652814", "GSM652815", "GSM652816", "GSM652817", "GSM652818", "GSM652819", "GSM652820", "GSM652821", "GSM652822", "GSM652823", "GSM652824", "GSM652825", "GSM652826", "GSM652827", "GSM652828", "GSM652829", "GSM652830", "GSM652831", "GSM652832", "GSM652833", "GSM652834", "GSM652835", "GSM652836", "GSM652837", "GSM652838", "GSM652839", "GSM652840", "GSM652841", "GSM652842", "GSM652843", "GSM652844", "GSM652845", "GSM652846"], "sample_count": 86, "group_sizes": null, "group_sizes_basis": null, "platform": "6244", "taxon": "Homo sapiens", "title": "Gene Expression Profiling Predicts the Development of Oral Cancer", "resolution_checked_at": "2026-11-01T18:06:43Z", "resolution_check": "identifier_resolution_only_no_download_no_licence_read", "downloaded": false, "licence_read": false}}, {"id": "data:geo:GSE41613", "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=GSE41613", "sha256": null, "bytes": null, "access": "open", "state": "resolved", "reader": null, "reader_status": "reader_required", "identifier": "GSE41613", "version": null, "note": "The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request.", "structure": {"repository": "GEO", "access_route": "GEO accession URL", "access_statement_verbatim": "The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request.", "access_statement_kind": "data_availability_statement", "sample_accessions": ["GSM1020099", "GSM1020100", "GSM1020101", "GSM1020102", "GSM1020103", "GSM1020104", "GSM1020105", "GSM1020106", "GSM1020107", "GSM1020108", "GSM1020109", "GSM1020110", "GSM1020111", "GSM1020112", "GSM1020113", "GSM1020114", "GSM1020115", "GSM1020116", "GSM1020117", "GSM1020118", "GSM1020119", "GSM1020120", "GSM1020121", "GSM1020122", "GSM1020123", "GSM1020124", "GSM1020125", "GSM1020126", "GSM1020127", "GSM1020128", "GSM1020129", "GSM1020130", "GSM1020131", "GSM1020132", "GSM1020133", "GSM1020134", "GSM1020135", "GSM1020136", "GSM1020137", "GSM1020138", "GSM1020139", "GSM1020140", "GSM1020141", "GSM1020142", "GSM1020143", "GSM1020144", "GSM1020145", "GSM1020146", "GSM1020147", "GSM1020148", "GSM1020149", "GSM1020150", "GSM1020151", "GSM1020152", "GSM1020153", "GSM1020154", "GSM1020155", "GSM1020156", "GSM1020157", "GSM1020158", "GSM1020159", "GSM1020160", "GSM1020161", "GSM1020162", "GSM1020163", "GSM1020164", "GSM1020165", "GSM1020166", "GSM1020167", "GSM1020168", "GSM1020169", "GSM1020170", "GSM1020171", "GSM1020172", "GSM1020173", "GSM1020174", "GSM1020175", "GSM1020176", "GSM1020177", "GSM1020178", "GSM1020179", "GSM1020180", "GSM1020181", "GSM1020182", "GSM1020183", "GSM1020184", "GSM1020185", "GSM1020186", "GSM1020187", "GSM1020188", "GSM1020189", "GSM1020190", "GSM1020191", "GSM1020192", "GSM1020193", "GSM1020194", "GSM1020195"], "sample_count": 97, "group_sizes": null, "group_sizes_basis": null, "platform": "570", "taxon": "Homo sapiens", "title": "A 13-gene signature prognostic of HPV-negative OSCC: discovery and external validation", "resolution_checked_at": "2026-11-01T15:39:54Z", "resolution_check": "identifier_resolution_only_no_download_no_licence_read", "downloaded": false, "licence_read": false}}, {"id": "data:geo:GSE9844", "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=GSE9844", "sha256": null, "bytes": null, "access": "open", "state": "resolved", "reader": null, "reader_status": "reader_required", "identifier": "GSE9844", "version": null, "note": "The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request.", "structure": {"repository": "GEO", "access_route": "GEO accession URL", "access_statement_verbatim": "The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request.", "access_statement_kind": "data_availability_statement", "sample_accessions": ["GSM248238", "GSM248650", "GSM248651", "GSM248652", "GSM248653", "GSM248655", "GSM248656", "GSM248658", "GSM248659", "GSM248660", "GSM248661", "GSM248662", "GSM248663", "GSM248664", "GSM248665", "GSM248667", "GSM248668", "GSM248669", "GSM248670", "GSM248671", "GSM248672", "GSM248673", "GSM248674", "GSM248675", "GSM248676", "GSM248677", "GSM248678", "GSM248679", "GSM248680", "GSM248681", "GSM248682", "GSM248683", "GSM248684", "GSM248685", "GSM248686", "GSM248687", "GSM248780", "GSM248781"], "sample_count": 38, "group_sizes": null, "group_sizes_basis": null, "platform": "570", "taxon": "Homo sapiens", "title": "Transcriptomic Dissection of Tongue Squamous Cell Carcinoma", "resolution_checked_at": "2026-11-01T15:39:55Z", "resolution_check": "identifier_resolution_only_no_download_no_licence_read", "downloaded": false, "licence_read": false}}], "gaps": ["asset:data:geo:GSE26549 not downloaded", "asset:data:geo:GSE26549 licence not read", "asset:data:geo:GSE41613 not downloaded", "asset:data:geo:GSE41613 licence not read", "asset:data:geo:GSE9844 not downloaded", "asset:data:geo:GSE9844 licence not read"]}, "analysis": {"state": "complete", "methods": [{"id": "method-1", "family": "survival_analysis", "specific": "cox_proportional_hazards", "label": "Survival analysis", "dsl_operations": ["cox_ph"], "missing_operations": [], "expressible": true, "parameters": {}, "data_shapes": ["time_to_event_with_censoring"], "evidence": {"score": 1.0, "band": "verbatim", "quote": "The extracted features from the autoencoder were examined through the univariate Cox proportional hazards (Cox-PH) model [30] to select significant features (P<0.05)."}}, {"id": "method-2", "family": "survival_analysis", "specific": "log_rank_test", "label": "Survival analysis", "dsl_operations": ["logrank_test"], "missing_operations": [], "expressible": true, "parameters": {"p_value": "groups were identified (P value of log-rank test = 0"}, "data_shapes": ["time_to_event_with_censoring"], "evidence": {"score": 0.7, "band": "medium", "quote": "log-rank"}}, {"id": "method-3", "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-4", "family": "clustering", "specific": "hierarchical", "label": "Clustering", "dsl_operations": ["hierarchical_cluster"], "missing_operations": [], "expressible": true, "parameters": {}, "data_shapes": ["feature_matrix"], "evidence": {"score": 0.7, "band": "medium", "quote": "hierarchical clustering"}}], "targets": [{"id": "target-published-estimate", "claim_id": "claim-target-result", "operation": "cox_ph", "expected_value": 0.003, "unit": "p-value", "tolerance": null, "state": "complete", "note": "The threshold the paper applied to its own result, as the paper states it. A run that reproduces the effect is judged against this threshold."}], "gaps": []}, "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": ["cox_ph"], "missing_operations": [], "basis": "The paper names the method it used and the corpus's method extractor maps it onto the DSL catalogue: cox_proportional_hazards."}, "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": ["GSE26549", "GSE9844", "GSE41613"], "resolved_accessions": ["GSE26549", "GSE9844", "GSE41613"], "statements": ["The datasets is publically available on https://www.ncbi.nlm.nih.gov/geo/. All analyzed during the current study are available from the corresponding author on request."], "basis": "The paper names GSE26549, GSE9844, GSE41613 in GEO repository and the accession resolves in the repository. No dataset was downloaded and no licence was read."}}, "can_run": {"dsl_operations": ["cox_ph"], "assets": ["source:pmid:36849997", "data:geo:GSE26549"], "steps": ["run_analysis"]}, "must_build": {"dsl_operations": [], "converters": [], "readers": [], "capabilities": [], "data_gaps": ["data:geo:GSE26549"]}, "gaps": []}, "review": {"state": "partial", "machine_critique": {"kind": "extraction_review", "reader": {"model": "MiniMax-M3.1-Flash", "provider": "minimax", "provider_family": "minimax", "read_at": "2026-11-01T18:32:19Z", "independence": "single_reader"}, "evidence_check": {"method": "verbatim", "availability": "verbatim"}, "calls": 5, "usage": {"prompt_tokens": 13325, "completion_tokens": 2039, "cached_tokens": 128}}, "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:36849997", "type": "paper", "label": "pmid:36849997"}, {"id": "source:pmid:36849997", "type": "manuscript", "label": "PMID 36849997 full text"}, {"id": "data:geo:GSE26549", "type": "dataset", "label": "GEO GSE26549"}], "activities": [{"id": "activity-reading-sci-2026-11-01", "type": "classification", "used": [{"id": "pmid:36849997", "type": "paper", "label": "pmid:36849997"}, {"id": "source:pmid:36849997", "type": "manuscript", "label": "PMID 36849997 full text"}, {"id": "data:geo:GSE26549", "type": "dataset", "label": "GEO GSE26549"}], "agent": "agent:corpus-assay-engine"}, {"id": "activity-read-36849997", "type": "extraction", "used": ["source:pmid:36849997"], "agent": "agent:minimax-reader"}, {"id": "activity-resolve-GSE26549", "type": "accession_resolution", "used": ["data:geo:GSE26549"], "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:36849997", "parent_hash": "sha256:8dd70216f87b3b5a901c28a65477eaca7d755cdbbacebdfaac1e7d3250c87356", "fork_of": null, "generation": 2}, "gaps": ["asset:data:geo:GSE26549 licence not read", "asset:data:geo:GSE26549 not downloaded", "asset:data:geo:GSE41613 licence not read", "asset:data:geo:GSE41613 not downloaded", "asset:data:geo:GSE9844 licence not read", "asset:data:geo:GSE9844 not downloaded", "execution.not_run", "review.human_critique", "review.independent_reader", "review.machine_critique"], "assays": [{"role": "analysis", "definition_id": "survival_analysis", "definition_version": "f2132e96ae6c7a4e", "definition_digest": null, "required_operations": ["cox_ph"], "support_state": "supported", "source_hashes": {"dsl_catalogue": "f2132e96ae6c7a4e"}, "confidence": 1.0, "evidence": ["The extracted features from the autoencoder were examined through the univariate Cox proportional hazards (Cox-PH) model [30] to select significant features (P<0.05)."], "order_basis": "analysis_only", "id": "assay:1"}]}