Integrative analysis and machine learning on cancer genomics data using the Cancer Systems Biology Database (CancerSysDB).

{"sci_version": "1.2", "schema": "https://corpus.bntrg.co.uk/sci/1.0/sci.schema.json", "id": "pmid:29699486", "kind": "science-record", "revision": 7, "content_hash": "sha256:221cac93c4e636ab1db60b96ef6075460882a0b24689a1cac056f77f437b5be7", "created_at": "2026-11-01T15:40:33Z", "updated_at": "2026-11-01T21:59:49Z", "generator": {"name": "corpus-assay-engine", "run_id": "reading-sci-2026-11-01"}, "paper": {"state": "complete", "title": "Integrative analysis and machine learning on cancer genomics data using the Cancer Systems Biology Database (CancerSysDB).", "doi": "10.1186/s12859-018-2157-7", "pmid": "29699486", "authors": ["Rasmus Krempel", "Pranav Kulkarni", "Annie Yim", "Ulrich Lang", "Bianca Habermann", "Peter Frommolt"], "journal": "BMC bioinformatics", "year": 2018, "abstract": "BACKGROUND: Recent cancer genome studies on many human cancer types have relied on multiple molecular high-throughput technologies. Given the vast amount of data that has been generated, there are surprisingly few databases which facilitate access to these data and make them available for flexible analysis queries in the broad research community. If used in their entirety and provided at a high structural level, these data can be directed into constantly increasing databases which bear an enormous potential to serve as a basis for machine learning technologies with the goal to support research and healthcare with predictions of clinically relevant traits.\nRESULTS: We have developed the Cancer Systems Biology Database (CancerSysDB), a resource for highly flexible queries and analysis of cancer-related data across multiple data types and multiple studies. The CancerSysDB can be adopted by any center for the organization of their locally acquired data and its integration with publicly available data from multiple studies. A publicly available main instance of the CancerSysDB can be used to obtain highly flexible queries across multiple data types as shown by highly relevant use cases. In addition, we demonstrate how the CancerSysDB can be used for predictive cancer classification based on whole-exome data from 9091 patients in The Cancer Genome Atlas (TCGA) research network.\nCONCLUSIONS: Our database bears the potential to be used for large-scale integrative queries and predictive analytics of clinically relevant traits.", "claims": [{"id": "claim-data-access", "statement": "The datasets analysed in the current study are available on the in the Genomic Data Commons (GDC) Data Portal at https://gdc.cancer.gov/.\nDatabase URL: https://cancersys.uni-koeln.de.\nSource code: https://github.com/RRZK/CancerSysDB.", "kind": "methodological", "target": null, "evidence": [{"asset": "source:pmid:29699486", "locator": "sections[15] \"Availability of data and materials\" paragraphs 1-4", "quote": "The datasets analysed in the current study are available on the in the Genomic Data Commons (GDC) Data Portal at https://gdc.cancer.gov/.\nDatabase URL: https://cancersys.uni-koeln.de.\nSource code: https://github.com/RRZK/CancerSysDB."}], "state": "complete"}, {"id": "claim-target-result", "statement": "The paper reports that a random forest trained on CancerSysDB somatic-mutation data from 9,091 TCGA patients classified 1,521 of 3,085 held-out patients correctly (49.3%), compared with 182 correct assignments (5.9%) for random guessing.", "kind": "qualitative", "target": null, "evidence": [{"asset": "source:pmid:29699486", "locator": "sections[8] \"Prediction of cancer types with random forests\" paragraph 3", "quote": "In the training phase, a random forest consisting of 1000 trees is trained on all data available in the CancerSysDB."}, {"asset": "source:pmid:29699486", "locator": "sections[5] \"Data preparation\" paragraph 2", "quote": "All TCGA data were obtained as level 3 data from the Legacy Archive of The Cancer Genome Atlas (TCGA) data portal. Data on somatic mutations were based on whole-exome sequencing with MAF files obtained from the Firehose pipeline of the Genome Data Analysis Center (GDAC) at the Broad Institute."}, {"asset": "source:pmid:29699486", "locator": "sections[8] \"Prediction of cancer types with random forests\" paragraph 4", "quote": "Out of these 3085 patients in the test set, 1521 (49.3%) were assigned to the correct class (Fig. 2), whereas a random guess of the class would have produced a correct class assignment in only 182 cases (5.9%)."}], "state": "complete"}], "gaps": []}, "experiment": {"state": "partial", "objective": "The paper reports that a random forest trained on CancerSysDB somatic-mutation data from 9,091 TCGA patients classified 1,521 of 3,085 held-out patients correctly (49.3%), compared with 182 correct assignments (5.9%) for random guessing.", "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:29699486"], "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": "blocked", "blocked_on": "data:accession_absent", "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": [], "outputs": ["the source data with its repository provenance"], "instruments": [], "reagents": []}, {"id": "run_analysis", "order": 3, "name": "Run the published analysis", "kind": "analysis", "state": "blocked", "blocked_on": "method:machine_learning_random_forest", "description": "Apply the paper's statistical method to the resolved series and compare the result with the published target.", "duration_hours": null, "parameters": {}, "inputs": [], "outputs": ["the replicated result and its comparison"], "instruments": [], "reagents": []}], "materials": [], "instruments": [], "reagents": [], "conditions": {}, "gaps": ["step:run_analysis blocked on method:machine_learning_random_forest", "step:obtain_source_data blocked on data:accession_absent"]}, "data": {"state": "partial", "assets": [{"id": "source:pmid:29699486", "role": "manuscript", "format": "xml", "media_type": "application/xml", "uri": "https://doi.org/10.1186/s12859-018-2157-7", "sha256": "sha256:a31f43048cefc9a9f397790eb01e349e2e59615b25b859fb3b3e9dd8901c4cfa", "bytes": 35785, "access": "open", "state": "resolved", "reader": null, "reader_status": "reader_required", "identifier": "10.1186/s12859-018-2157-7", "version": null, "note": "Full article fetched from PMC; the checksum covers the rendered section body, not the original XML bytes.", "structure": {"pmcid": "PMC5921751", "source": "europe_pmc_fulltext", "coverage": "whole_body", "text_hash_scope": "rendered_section_body"}}, {"id": "asset:table:pmid:29699486:1", "role": "table", "format": "jats-table", "media_type": "application/vnd.jats.table+json", "uri": null, "sha256": "sha256:f0c456c47a18c87da3d17e1fc48a1a318a56704d9f47d3b2089eea9865b2ffd2", "bytes": 911, "access": "open", "state": "verified", "reader": "read_table", "reader_status": "available", "identifier": "section[3]/table-wrap[1]#Tab1", "version": null, "note": "Comparison of various features of the CancerSysDB with those of other cancer genomics data integration tools", "structure": {"pmcid": "PMC5921751", "section_index": 3, "section_title": "Implementation", "row_count": 7, "column_count": 5}, "table": {"label": "Table 1", "caption": "Comparison of various features of the CancerSysDB with those of other cancer genomics data integration tools", "rows": [["", "CancerSysDB", "TCGAbiolinks", "RTCGA", "cBio portal"], ["GUI", "Web framework based on Groovy/Grails", "Based on Shiny", "None", "Web framework based on Spring Java"], ["Query schema", "Hibernate", "R scripting", "R scripting", "SQL"], ["Data upload", "Parametrized CSV file upload", "Direct access to GDC through API", "Data packages available on Bioconductor", "CSV files plus meta file"], ["Query definition", "JSON-based", "Combination of R commands", "Combination of R commands", "REST-based API"], ["Portability", "Native Docker implementation", "Hosted on Bioconductor", "Hosted on Bioconductor", "Hosted on GitHub"], ["Programming skills required", "No", "Yes", "Yes", "No"]], "row_count": 7, "column_count": 5, "numeric_cells": 0, "header_rows": 1, "locator": "section[3]/table-wrap[1]#Tab1"}}, {"id": "asset:table:pmid:29699486:2", "role": "table", "format": "jats-table", "media_type": "application/vnd.jats.table+json", "uri": null, "sha256": "sha256:29c52585c1d60da8a361d90aa17f7ed7aed81ef55521f80976fe2bdf345e8eeb", "bytes": 1123, "access": "open", "state": "verified", "reader": "read_table", "reader_status": "available", "identifier": "section[7]/table-wrap[1]#Tab2", "version": null, "note": "Results of TP53-dependent analysis of genomic and clinical characteristics", "structure": {"pmcid": "PMC5921751", "section_index": 7, "section_title": "TP53-dependent analysis of overall survival, genome stability, and mutation types", "row_count": 17, "column_count": 8}, "table": {"label": "Table 2", "caption": "Results of TP53-dependent analysis of genomic and clinical characteristics", "rows": [["(a)"], ["", "Patients", "Events", "5-year survival rate [%]", "Median survival", "95% CI"], ["TP53 mutant", "3772", "1237", "47.4", "1670", "[1526; 1818]"], ["TP53 non-mutant", "5672", "1128", "66.9", "3736", "[3262; 4267]"], ["(b)"], ["", "Patients", "CNAs [Mb]"], ["TP53 mutant", "133", "74.5"], ["TP53 non-mutant", "246", "50.5"], ["(c)"], ["", "", "TP53", "ATM"], ["VarType", "All", "Mutant [%] (n = 320)", "Non-mutant [%] (n = 265)", "p-value", "Mutant [%] (n = 49)", "Non-mutant [%] (n = 536)", "p-value"], ["A > C or T > G", "3.5", "3.3", "3.6", "< 0.0001", "3.9", "3.4", "0.2160"], ["A > G or T > C", "9.9", "9.2", "10.7", "< 0.0001", "9.6", "9.9", "0.7695"], ["A > T or T > A", "8.1", "8.4", "7.8", "0.0005", "8.6", "8.1", "0.4584"], ["C > G or G > C", "13.6", "13.9", "13.1", "< 0.0001", "13.2", "13.6", "0.3790"], ["C > T or G > A", "32.7", "30.0", "36.0", "< 0.0001", "28.6", "33.0", "0.5121"], ["G > T or C > A", "32.3", "35.2", "28.8", "0.0001", "36.0", "32.0", "0.4940"]], "row_count": 17, "column_count": 8, "numeric_cells": 67, "header_rows": 0, "locator": "section[7]/table-wrap[1]#Tab2"}}, {"id": "asset:table:pmid:29699486:3", "role": "table", "format": "jats-table", "media_type": "application/vnd.jats.table+json", "uri": null, "sha256": "sha256:b8db4be4bfdb918113a98b48d88a85347aca700e9f74c6197918285c579a419b", "bytes": 1830, "access": "open", "state": "verified", "reader": "read_table", "reader_status": "available", "identifier": "section[8]/table-wrap[1]#Tab3", "version": null, "note": "Classes of carcinomas used for random forest prediction of cancer types", "structure": {"pmcid": "PMC5921751", "section_index": 8, "section_title": "Prediction of cancer types with random forests", "row_count": 32, "column_count": 5}, "table": {"label": "Table 3", "caption": "Classes of carcinomas used for random forest prediction of cancer types", "rows": [["Class name", "TCGA cohorts", "Sample size"], ["Total", "Training set", "Test set"], ["Adrenal gland", "Adrenocortical carcinoma (ACC)", "271", "179", "92"], ["Pheochromocytoma and paraganglioma (PCPG)"], ["Bladder", "Urothelial carcinoma (BLCA)", "411", "272", "139"], ["Brain", "Lower grade glioma (LGG)", "515", "340", "175"], ["Breast", "Breast invasive carcinoma (BRCA)", "1077", "711", "366"], ["Gastrointestinal", "Esophageal carcinoma (ESCA)", "1237", "817", "420"], ["Stomach adenocarcinoma (STAD)"], ["Colon adenocarcinoma (COAD)"], ["Rectum adenocarcinoma (READ)"], ["Cholangiocarcinoma (CHOL)"], ["Head & Neck", "Head and neck squamous cell carcinoma (HNSC)", "590", "390", "200"], ["Uveal melanoma (UVM)"], ["Hematologic", "Acute myeloid leukemia (LAML)", "321", "212", "109"], ["Diffuse large B-cell lymphoma (DLBC)"], ["Thymoma (THYM)"], ["Kidney", "Kidney Chromophobe (KICH)", "738", "488", "250"], ["Renal clear cell carcinoma (KIRC)"], ["Renal papillary cell carcinoma (KIRP)"], ["Liver", "Hepatocellular carcinoma (LIHC)", "321", "212", "109"], ["Ovary", "Ovarian serous cystadenocatcinoma (OV)", "437", "289", "148"], ["Pancreas", "Pancreatic adenocarcinoma (PAAD)", "184", "122", "62"], ["Prostate", "Prostate adenocarcinoma (PRAD)", "498", "329", "169"], ["Skin", "Cutaneous melanoma (SKCM)", "104", "69", "35"], ["Testis", "Testicular germ cell tumors (TGCT)", "150", "99", "51"], ["Thoracic", "Lung adenocarcinoma (LUAD)", "1143", "755", "388"], ["Lung squamous cell carcinoma (LUSC)"], ["Mesothelioma (MESO)"], ["Thyroid", "Thyroid carcinoma (THCA)", "496", "327", "169"], ["Uterus", "Uterine carcinosarcoma (UCS)", "598", "395", "203"], ["Uterine corpus endometrial carcinoma (UCEC)"]], "row_count": 32, "column_count": 5, "numeric_cells": 51, "header_rows": 2, "locator": "section[8]/table-wrap[1]#Tab3"}}, {"id": "asset:table:pmid:29699486:4", "role": "table", "format": "jats-table", "media_type": "application/vnd.jats.table+json", "uri": null, "sha256": "sha256:3174b217d2e40be55ab57bbd4273ec906856448a795156b9dfc1029f4637c033", "bytes": 623, "access": "open", "state": "verified", "reader": "read_table", "reader_status": "available", "identifier": "section[9]/table-wrap[1]#Tab4", "version": null, "note": "Averaged log2-fold changes of SUCLG1 and SUCLG2 mRNAs in different tumor stages of KIRP cancer patients", "structure": {"pmcid": "PMC5921751", "section_index": 9, "section_title": "Analyzing TCA-cycle genes in kidney renal papillary cell carcinoma (KIRP)", "row_count": 6, "column_count": 8}, "table": {"label": "Table 4", "caption": "Averaged log2-fold changes of SUCLG1 and SUCLG2 mRNAs in different tumor stages of KIRP cancer patients", "rows": [["Stage", "# Patients", "Female/Male", "Alive/Dead", "SUCLG1", "SUCLG2"], ["log2 FC", "p-value", "log2 FC", "p-value"], ["I", "15", "5 / 10", "13 / 2", "−0.473", "0.132", "−0.338", "0.307"], ["II", "1", "0 / 1", "1 / 0", "−1.163", "0.082", "0.137", "0.431"], ["III", "11", "5 / 6", "8 / 3", "−0.835", "0.018", "−0.760", "0.028"], ["IV", "4", "0 / 4", "3 / 1", "−1.975", "0.066", "−1.664", "0.054"]], "row_count": 6, "column_count": 8, "numeric_cells": 28, "header_rows": 2, "locator": "section[9]/table-wrap[1]#Tab4"}}], "gaps": []}, "analysis": {"state": "partial", "methods": [{"id": "method-1", "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": 0.8, "band": "high", "quote": "random forest"}}], "targets": [], "gaps": ["method.machine_learning_random_forest"]}, "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": "blocked", "tier": "R4", "blocking_axis": "data", "blocked_on": "data:accession_absent", "axes": {"method": {"tier": "R1", "expressible": false, "required_operations": [], "missing_operations": ["machine_learning_random_forest"], "basis": "The paper names the method it used and the corpus's method extractor maps it onto the DSL catalogue: random_forest."}, "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": "R4", "reachable": false, "accessions": [], "resolved_accessions": [], "statements": ["The datasets analysed in the current study are available on the in the Genomic Data Commons (GDC) Data Portal at https://gdc.cancer.gov/.\nDatabase URL: https://cancersys.uni-koeln.de.\nSource code: https://github.com/RRZK/CancerSysDB."], "basis": "The paper names no accession to its source data, so the data has to come from the authors."}}, "can_run": {"dsl_operations": [], "assets": ["source:pmid:29699486", "asset:table:pmid:29699486:1", "asset:table:pmid:29699486:2", "asset:table:pmid:29699486:3", "asset:table:pmid:29699486:4"], "steps": []}, "must_build": {"dsl_operations": [{"operation": "machine_learning_random_forest", "unlocks_papers": null, "priority": null}], "converters": [], "readers": [], "capabilities": [], "data_gaps": ["data:geo:None"]}, "gaps": ["data:accession_absent"]}, "review": {"state": "partial", "machine_critique": {"kind": "extraction_review", "reader": {"model": "MiniMax-M3.1-Flash", "provider": "minimax", "provider_family": "minimax", "read_at": "2026-10-31T18:34:19Z", "independence": "same_provider"}, "evidence_check": {"method": null, "availability": "verbatim"}, "calls": 8, "usage": {"prompt_tokens": 69543, "completion_tokens": 4563, "cached_tokens": 1832}}, "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:29699486", "type": "paper", "label": "pmid:29699486"}, {"id": "source:pmid:29699486", "type": "manuscript", "label": "PMID 29699486 full text"}, {"id": "data:geo:None", "type": "dataset", "label": "GEO None"}], "activities": [{"id": "activity-reading-sci-2026-11-01", "type": "classification", "used": [{"id": "pmid:29699486", "type": "paper", "label": "pmid:29699486"}, {"id": "source:pmid:29699486", "type": "manuscript", "label": "PMID 29699486 full text"}, {"id": "data:geo:None", "type": "dataset", "label": "GEO None"}], "agent": "agent:corpus-assay-engine"}, {"id": "activity-read-29699486", "type": "extraction", "used": ["source:pmid:29699486"], "agent": "agent:minimax-reader"}, {"id": "activity-resolve-None", "type": "accession_resolution", "used": ["data:geo:None"], "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:29699486", "parent_hash": "sha256:4004f24d3e137f180871fbc824f295602e53d1664e2d5e76277014eb747fe595", "fork_of": null, "generation": 6}, "gaps": ["execution.not_run", "method.machine_learning_random_forest", "review.human_critique", "review.independent_reader", "review.machine_critique", "step:obtain_source_data blocked on data:accession_absent", "step:run_analysis blocked on method:machine_learning_random_forest"], "assays": [{"role": "data_generation", "definition_id": "apoptosis_assay", "definition_version": "99e1195c7d6f", "definition_digest": null, "required_operations": [], "support_state": "partial", "source_hashes": {"assay_lexicon": "99e1195c7d6f"}, "confidence": 0.93, "evidence": ["s likely caused by the impaired induction of apoptosis upon these exogenic damages.Genomic complexi", "ributed to the role of TP53 as a mediator of apoptosis in response to somatically acquired DNA dama"], "order_basis": "paper_text_order:2882", "id": "assay:1"}, {"role": "analysis", "definition_id": "machine_learning", "definition_version": "f2132e96ae6c7a4e", "definition_digest": null, "required_operations": [], "support_state": "partial", "source_hashes": {"dsl_catalogue": "f2132e96ae6c7a4e"}, "confidence": 0.8, "evidence": ["random forest"], "order_basis": "consumes_earlier_output", "id": "assay:2"}]}