Prediction of molecular subtypes for endometrial cancer based on hierarchical foundation model.

{"sci_version": "1.0", "schema": "https://corpus.bntrg.co.uk/sci/1.0/sci.schema.json", "id": "pmid:39932017", "kind": "science-record", "revision": 2, "content_hash": "sha256:085aa9d683773e4682e02dd914fb0c5b7797985767b3d04f1c6065b41a5cbe71", "created_at": "2026-11-01T10:25:49Z", "updated_at": "2026-11-01T14:01:34Z", "generator": {"name": "corpus-assay-engine", "converter": "sep_to_sci_v1", "source_package_schema": "urn:bntrg:sep:1.0.0"}, "paper": {"state": "partial", "title": "Prediction of molecular subtypes for endometrial cancer based on hierarchical foundation model.", "doi": "10.1093/bioinformatics/btaf059", "pmid": "39932017", "authors": null, "journal": "Bioinformatics (Oxford, England)", "year": 2025, "abstract": null, "claims": [{"id": "claim-1", "statement": "the headline numerical result the paper states for its main finding", "kind": "quantitative", "state": "complete", "target": {"value": 0.879, "unit": "macro-average auroc", "tolerance": null, "source": "chars 999-1107 of source sha256 7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7"}, "evidence": [{"asset": "source:pmid:39932017", "locator": "chars 999-1107 of source sha256 7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7", "quote": "Our model demonstrates a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation."}]}], "gaps": ["abstract", "authors"]}, "experiment": {"state": "partial", "objective": "the headline numerical result the paper states for its main finding", "steps": [{"id": "read_published_estimate", "order": 1, "name": "Read the published estimate and its interval from the source", "kind": "other", "state": "complete", "blocked_on": null, "description": "The number a replication must land on, and the sentence it came from.", "duration_hours": null, "parameters": {}, "inputs": ["source:pmid:39932017"], "outputs": ["the published estimate, its unit and its interval"], "instruments": [], "reagents": []}, {"id": "obtain_source_data", "order": 2, "name": "Obtain the source data behind the published result", "kind": "data-preparation", "state": "blocked", "blocked_on": "data:accession_absent", "description": "The published result cannot be recomputed without the data the paper analysed.", "duration_hours": null, "parameters": {}, "inputs": [], "outputs": ["the source data with its provenance"], "instruments": [], "reagents": []}, {"id": "run_analysis", "order": 3, "name": "Run the analysis programme", "kind": "analysis", "state": "blocked", "blocked_on": "no owned computational capability is bound to this experiment", "description": "Re-run the paper's analysis on the source data and produce the registered result. | SEP input not carried into the asset list: materials:the source data behind the published result in pmid:39932017", "duration_hours": null, "parameters": {}, "inputs": [], "outputs": ["the registered result"], "instruments": [], "reagents": []}], "materials": [], "instruments": [], "reagents": [], "conditions": {}, "gaps": ["step:obtain_source_data blocked on data:accession_absent", "step:run_analysis blocked on no owned computational capability is bound to this experiment", "step:run_analysis input 'materials:the source data behind the published result in pmid:39932017' does not resolve to a declared asset"]}, "data": {"state": "partial", "assets": [{"id": "source:pmid:39932017", "role": "manuscript", "format": "xml", "media_type": "application/xml", "uri": "https://doi.org/10.1093/bioinformatics/btaf059", "sha256": "sha256:7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7", "bytes": null, "access": "open", "state": "resolved", "reader": null, "reader_status": "reader_required", "identifier": "10.1093/bioinformatics/btaf059", "version": null, "note": "Source read by the package builder: full_text.", "structure": {"access_state": "full_text"}}, {"id": "asset:table:pmid:39932017:1", "role": "table", "format": "jats-table", "media_type": "application/vnd.jats.table+json", "uri": null, "sha256": "sha256:22dbaffc1c14b444413f19f8def6eafb3f7cab5560032c081250b687a6a53f5c", "bytes": 1874, "access": "open", "state": "verified", "reader": "read_table", "reader_status": "available", "identifier": "section[11]/table-wrap[1]#btaf059-T1", "version": null, "note": "Five-fold cross-validation macro-average and class-wise AUROC comparison of our method and MIL methods with ImageNet pretrained ResNet or UNI as the feature extraction network.a", "structure": {"pmcid": "PMC11878776", "section_index": 11, "section_title": "3.3 Comparison of proposed method and MIL with different feature extraction networks", "row_count": 12, "column_count": 6}, "table": {"label": "Table 1.", "caption": "Five-fold cross-validation macro-average and class-wise AUROC comparison of our method and MIL methods with ImageNet pretrained ResNet or UNI as the feature extraction network.a", "rows": [["Methods", "MMRd", "NSMP", "p53abn", "POLE mut", "Macro"], ["CLAM-SBIPR", "0.559 (0.499–0.619)", "0.689 (0.560–0.817)", "0.680 (0.449–0.910)", "0.607 (0.338–0.875)", "0.634 (0.521–0.746)"], ["Attention-MILIPR", "0.576 (0.540–0.611)", "0.717 (0.570–0.864)", "0.695 (0.573–0.818)", "0.565 (0.478–0.652)", "0.638 (0.572–0.705)"], ["SETMILIPR", "0.664 (0.628–0.700)", "0.730 (0.702–0.758)", "0.778 (0.720–0.835)", "0.631 (0.581–0.680)", "0.700 (0.678–0.723)"], ["DTFDIPR", "0.629 (0.578–0.680)", "0.747 (0.704–0.790)", "0.764 (0.694–0.833)", "0.633 (0.533–0.734)", "0.693 (0.647–0.739)"], ["TransMILIPR", "0.689 (0.612–0.766)", "0.783 (0.752–0.815)", "0.823 (0.742–0.905)", "0.685 (0.664–0.706)", "0.745 (0.719–0.772)"], ["CLAM-SBUNI", "0.648 (0.523–0.773)", "0.818 (0.730–0.906)", "0.820 (0.670–0.969)", "0.628 (0.415–0.841)", "0.729 (0.693–0.764)"], ["Attention-MILUNI", "0.686 (0.664–0.707)", "0.714 (0.662–0.767)", "0.820 (0.698–0.942)", "0.581 (0.482–0.681)", "0.700 (0.664–0.736)"], ["SETMILUNI", "0.765 (0.696–0.835)", "0.812 (0.774–0.849)", "0.852 (0.792–0.912)", "0.773 (0.732–0.815)", "0.801 (0.760–0.841)"], ["DTFDUNI", "0.788 (0.757–0.820)", "0.864 (0.840–0.889)", "0.900 (0.856–0.943)", "0.761 (0.729–0.794)", "0.828 (0.806–0.851)"], ["TransMILUNI", "0.785 (0.735–0.834)", "0.853 (0.817–0.889)", "0.906 (0.833–0.978)", "0.808 (0.773–0.842)", "0.838 (0.805–0.871)"], ["Ours", "0.829 (0.816–0.843)", "0.899 (0.867–0.931)", "0.899 (0.836–0.962)", "0.886 (0.853–0.919)", "0.879 (0.853–0.904)"]], "row_count": 12, "column_count": 6, "numeric_cells": 56, "header_rows": 1, "locator": "section[11]/table-wrap[1]#btaf059-T1"}}], "gaps": ["asset:source:pmid:39932017 not verified"]}, "analysis": {"state": "absent", "methods": [], "targets": [], "gaps": ["method", "targets"]}, "capability": {"state": "absent", "requirements": [], "gaps": ["capability"]}, "readiness": {"state": "partial", "tier": null, "blocking_axis": null, "blocked_on": null, "axes": {}, "can_run": {"dsl_operations": [], "assets": ["source:pmid:39932017", "asset:table:pmid:39932017:1"], "steps": ["read_published_estimate"]}, "must_build": {"dsl_operations": [], "converters": [], "readers": ["xml"], "capabilities": [], "data_gaps": ["source:pmid:39932017"]}, "gaps": ["corpus_record_absent"]}, "review": {"state": "partial", "machine_critique": {"status": "passed", "providers": [{"provider": "minimax", "model": "MiniMax-M3.1-Flash", "prompt_version": "full-text-reading-v1", "prompt_hash": "ba02bcf6f15964b3d5510eacb2ba13ee651f4047505e5cccb4b629cd762d117a", "source_hash": "7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7", "raw_response_sha256": "sha256:c885622c70af732a5d10957b6f74fdce86588db060247dac582ae1819a70cb88", "raw_response_chars": 3878}, {"provider": "minimax_alt", "model": "MiniMax-M3", "prompt_version": "full-text-reading-v1", "prompt_hash": "e38a723da8fbd74ecb5e249fd06024f3a2c9c6857a4a213549c8d76fd2fc1292", "source_hash": "7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7", "raw_response_sha256": "sha256:6abe1316feee351a476dbf51fbc48a831622eb10eedb0ddb0ffd49fa4212fc6b", "raw_response_chars": 29755}], "target_providers": [{"provider": "minimax", "model": "MiniMax-M3.1-Flash", "prompt_version": "replication-target-v2", "prompt_hash": "0622670056191e8fcf368794b736ce8fb0eabd949f0d5caa988ac443abca0485", "source_hash": "7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7", "raw_response": "{\"parent_value\":0.879,\"unit\":\"macro-average AUROC\",\"evidence\":\"Our model demonstrates a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation.\",\"target_result\":\"The proposed model achieved a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in five-fold cross-validation.\",\"number_stated\":true}", "reading": {"parent_value": 0.879, "unit": "macro-average AUROC", "evidence": "Our model demonstrates a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation.", "target_result": "The proposed model achieved a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in five-fold cross-validation.", "number_stated": true, "source_locator": "chars 999-1107 of source sha256 7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7"}, "recorded_at": "2026-10-31T15:23:47.100434+00:00"}, {"provider": "minimax_alt", "model": "MiniMax-M3", "prompt_version": "replication-target-v2", "prompt_hash": "acdd98269e3baf619e624bab177d7fa6269f555c763615b6687363ffacbba259", "source_hash": "7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7", "raw_response": "<think>We need answer JSON only. Need identify headline numerical result main finding: macro-average AUROC 0.879. Unit paper prints likely \"macro-average AUROC\" or \"AUROC\"? Asked value paper states registered target. parent_value number 0.879, unit \"macro-average AUROC\" perhaps unit should be printed unit. Evidence exact contiguous quotation containing number and attribution. \"Our model demonstrates a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation.\" Exact. target_result perhaps \"macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation\" but headline number only? The JSON has target_result. Need likely concise. number_stated true.\n\nNeed ensure valid JSON en dash okay. Unit might be \"AUROC\" rather than \"macro-average AUROC\". User says unit must be unit paper prints. It prints \"macro-average AUROC of 0.879\". Could set unit \"macro-average AUROC\". Evidence exact quote. target_result perhaps \"The model achieved a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in five-fold cross-validation.\" But asks \"Report the headline number only\" likely don't clutter with CI? Yet target_result can state value. Could simply \"0.879 macro-average AUROC\". Evidence quote whole clause. Need not mention calculated. Let's formulate.\n\nExact source sentence:\n\"Our model demonstrates a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation.\"\nNo issue. JSON.</think>\n\n{\"parent_value\":0.879,\"unit\":\"macro-average AUROC\",\"evidence\":\"Our model demonstrates a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation.\",\"target_result\":\"0.879 macro-average AUROC\",\"number_stated\":true}", "reading": {"parent_value": 0.879, "unit": "macro-average AUROC", "evidence": "Our model demonstrates a macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation.", "target_result": "0.879 macro-average AUROC", "number_stated": true, "source_locator": "chars 999-1107 of source sha256 7728c9989532848d5760645b1af7cf6351f677a898572a66fae308a6e1a1b3d7"}, "recorded_at": "2026-10-31T15:23:42.827044+00:00"}], "agreement": true, "disagreements": {"target_result": {"minimax": "Macro-average AUROC of 0.879 (95% CI, 0.853–0.904) for predicting endometrial cancer molecular subtypes in five-fold cross-validation.", "minimax_alt": "macro-average AUROC of 0.879 (95% CI, 0.853–0.904) in a five-fold cross-validation", "decisive": false}, "physical_methods": {"minimax": ["Hematoxylin and eosin-stained whole slide image analysis"], "minimax_alt": [], "decisive": false}, "analysis_methods": {"minimax": ["Hierarchical foundation model fine-tuned from the UNI computational pathology foundation model", "Weakly supervised learning", "Five-fold cross-validation", "Macro-average area under the receiver operating characteristic curve (AUROC)"], "minimax_alt": ["hierarchical foundation model fine-tuned from UNI with weakly supervised learning", "five-fold cross-validation with 80% training and 20% testing in each fold", "macro-average and class-wise AUROC", "Accuracy, Precision, Recall, F1 scores, Sensitivity, Specificity, and NPV", "t-SNE visualization of extracted feature vectors"], "decisive": false}}, "agreement_basis": {"policy": "decisive_fields", "decisive_fields": ["work_type", "number_stated", "parent_value", "unit"]}, "blinded": true}, "human_critique": {"status": "not_run", "reviewer": null, "signed_at": null, "conflict_checked": false, "commentary": null}, "stages": [{"key": "target_materials", "status": "pending", "reason": null, "evidence": null}, {"key": "independent_readings", "status": "passed", "reason": "The independent readings are incomplete or disagree; the source evidence must resolve the difference.", "evidence": {"provider_errors": []}}, {"key": "source_resolution", "status": "pending", "reason": null, "evidence": null}, {"key": "protocol_reconstruction", "status": "pending", "reason": null, "evidence": null}, {"key": "control_execution", "status": "pending", "reason": null, "evidence": null}, {"key": "control_comparison", "status": "pending", "reason": null, "evidence": null}, {"key": "publication", "status": "pending", "reason": null, "evidence": null}], "gaps": ["review.human_critique", "review.machine_critique.passed"]}, "execution": {"state": "absent", "status": "not_run", "mode": null, "runs": [], "gaps": ["execution.not_run"]}, "provenance": {"entities": [{"id": "package:ddb1aa58c8e4f2ff35325e9a11e117b4ae5b961774eb1d6e229a1b416999fce3", "type": "sep_package", "label": "SEP v1 package for pmid:39932017"}], "activities": [{"id": "activity:conversion:pmid:39932017", "type": "conversion", "used": ["package:ddb1aa58c8e4f2ff35325e9a11e117b4ae5b961774eb1d6e229a1b416999fce3"], "agent": "agent:corpus-assay-engine"}], "agents": [{"id": "agent:minimax", "type": "model", "label": "minimax"}, {"id": "agent:minimax_alt", "type": "model", "label": "minimax_alt"}, {"id": "agent:corpus-assay-engine", "type": "software", "label": "Corpus assay engine"}], "source_run_id": "run-20261023T122449232913", "dsl_catalogue_fingerprint": "343fb30756ba6ec3", "capability_table_fingerprint": "2cb3e42f047d2ecf", "snapshot_id": "ba34b49490d9bfaf"}, "lineage": {"parent_id": "pmid:39932017", "parent_hash": "sha256:02a36daa6c281a7c6d61d06baad2e43494e84fa75982cdfcc1b7c1ee1304ba9a", "fork_of": null, "generation": 2}, "gaps": ["abstract", "asset:source:pmid:39932017 not verified", "authors", "capability", "corpus_record_absent", "execution.not_run", "method", "review.human_critique", "review.machine_critique.passed", "step:obtain_source_data blocked on data:accession_absent", "step:run_analysis blocked on no owned computational capability is bound to this experiment", "step:run_analysis input 'materials:the source data behind the published result in pmid:39932017' does not resolve to a declared asset", "targets"]}