{ "schema_version": 1, "research_line": "protein_response_retrieval", "status": "exploratory exact-condition retrieval within one published study; no independent therapeutic validation", "freeze": { "commit": "012e2c5", "frozen_utc": "2026-10-06T12:09:06Z", "before_scoring": true, "completed_utc": "2026-10-06T12:09:26Z" }, "experimental_context": { "cell_model": "MCF7", "treatment_hours": 12, "biological_replicates": 2, "measurement": "protein abundance, not activity or binding", "target": "same treatment condition in the opposite biological replicate" }, "counts": { "conditions": 78, "accessions": 15170, "gene_annotations": 8889, "observed_values": 998650, "missing_values": 1367870 }, "method": { "score": "cosine on pairwise observed log2 treatment/DMSO ratios; no centering or imputation", "primary_min_shared_accessions": 100, "gene_query_min_shared": 20, "unit_assumption": "Positive S2 values are interpreted as treatment/DMSO ratios from article method 2.5; workbook caption does not declare units.", "mask_control": "Jaccard similarity of observed accession sets", "value_control": "One fixed shuffle of observed query values, preserving missing positions; not a permutation test or p-value." }, "primary": { "observed_cosine": { "summary": { "queries": 156, "scored_queries": 156, "target_eligible_queries": 156, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 109, "top5_hits": 149, "top1": 0.6987179487179487, "top5": 0.9551282051282052, "mrr": 0.7999137436637437, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "by_direction": { "R1_to_R2": { "queries": 78, "scored_queries": 78, "target_eligible_queries": 78, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 53, "top5_hits": 73, "top1": 0.6794871794871795, "top5": 0.9358974358974359, "mrr": 0.779528342028342, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "R2_to_R1": { "queries": 78, "scored_queries": 78, "target_eligible_queries": 78, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 56, "top5_hits": 76, "top1": 0.717948717948718, "top5": 0.9743589743589743, "mrr": 0.8202991452991453, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } } }, "by_period": { "1P": { "queries": 50, "scored_queries": 50, "target_eligible_queries": 50, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 34, "top5_hits": 48, "top1": 0.68, "top5": 0.96, "mrr": 0.7986753246753246, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "2P": { "queries": 38, "scored_queries": 38, "target_eligible_queries": 38, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 20, "top5_hits": 35, "top1": 0.5263157894736842, "top5": 0.9210526315789473, "mrr": 0.6679824561403509, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "3P": { "queries": 68, "scored_queries": 68, "target_eligible_queries": 68, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 55, "top5_hits": 66, "top1": 0.8088235294117647, "top5": 0.9705882352941176, "mrr": 0.8745506535947712, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } } } }, "mask_jaccard": { "summary": { "queries": 156, "scored_queries": 156, "target_eligible_queries": 156, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 26, "top5_hits": 118, "top1": 0.16666666666666666, "top5": 0.7564102564102564, "mrr": 0.38629886851040696, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "by_direction": { "R1_to_R2": { "queries": 78, "scored_queries": 78, "target_eligible_queries": 78, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 13, "top5_hits": 60, "top1": 0.16666666666666666, "top5": 0.7692307692307693, "mrr": 0.3877286175363099, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "R2_to_R1": { "queries": 78, "scored_queries": 78, "target_eligible_queries": 78, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 13, "top5_hits": 58, "top1": 0.16666666666666666, "top5": 0.7435897435897436, "mrr": 0.3848691194845041, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } } }, "by_period": { "1P": { "queries": 50, "scored_queries": 50, "target_eligible_queries": 50, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 9, "top5_hits": 40, "top1": 0.18, "top5": 0.8, "mrr": 0.40203124653124656, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "2P": { "queries": 38, "scored_queries": 38, "target_eligible_queries": 38, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 5, "top5_hits": 20, "top1": 0.13157894736842105, "top5": 0.5263157894736842, "mrr": 0.3191096184517237, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "3P": { "queries": 68, "scored_queries": 68, "target_eligible_queries": 68, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 12, "top5_hits": 58, "top1": 0.17647058823529413, "top5": 0.8529411764705882, "mrr": 0.41227787735140675, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } } } }, "value_permuted_cosine": { "summary": { "queries": 156, "scored_queries": 156, "target_eligible_queries": 156, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 1, "top5_hits": 10, "top1": 0.00641025641025641, "top5": 0.0641025641025641, "mrr": 0.061486821296639856, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "by_direction": { "R1_to_R2": { "queries": 78, "scored_queries": 78, "target_eligible_queries": 78, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 0, "top5_hits": 7, "top1": 0.0, "top5": 0.08974358974358974, "mrr": 0.06138100440710032, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "R2_to_R1": { "queries": 78, "scored_queries": 78, "target_eligible_queries": 78, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 1, "top5_hits": 3, "top1": 0.01282051282051282, "top5": 0.038461538461538464, "mrr": 0.061592638186179376, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } } }, "by_period": { "1P": { "queries": 50, "scored_queries": 50, "target_eligible_queries": 50, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 1, "top5_hits": 5, "top1": 0.02, "top5": 0.1, "mrr": 0.09083720269170989, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "2P": { "queries": 38, "scored_queries": 38, "target_eligible_queries": 38, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 0, "top5_hits": 3, "top1": 0.0, "top5": 0.07894736842105263, "mrr": 0.0465364813088095, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } }, "3P": { "queries": 68, "scored_queries": 68, "target_eligible_queries": 68, "target_eligible_fraction": 1.0, "candidate_count_min": 78, "candidate_count_max": 78, "top1_hits": 0, "top5_hits": 2, "top1": 0.0, "top5": 0.029411764705882353, "mrr": 0.04826026026405237, "uniform_rank_reference": { "top1": 0.01282051282051282, "top5": 0.0641025641025641, "mrr": 0.06333744937304378 } } } } }, "secondary": { "status": "secondary curated literature signatures; selection and study dependence limit generalization", "summary": { "queries": 11, "scored_queries": 11, "target_eligible_queries": 11, "target_eligible_fraction": 1.0, "candidate_count_min": 44, "candidate_count_max": 78, "top1_hits": 6, "top5_hits": 7, "top1": 0.5454545454545454, "top5": 0.6363636363636364, "mrr": 0.6073921028466484, "uniform_rank_reference": { "top1": 0.013721127357490993, "top5": 0.06860563678745496, "mrr": 0.06661405706088505 } }, "coverage": { "source_signatures": 20, "searched_signatures": 20, "mapped_for_identity_accuracy": 11, "excluded_from_identity_accuracy": 9, "with_any_candidate": 19, "with_no_candidate": 1, "candidate_count_min": 0, "candidate_count_max": 78 }, "note": "The source authors selected and previously used these literature signatures. The 11 mapped signatures alone form the identity-accuracy denominator; all 20 were searched." }, "timings": { "parse_seconds": 6.497559834271669, "computation_seconds": 1.5578866661526263, "excludes": "download, plotting and JSON saving" }, "verification": { "tests_passed": 41, "tests_skipped": 0, "scope": "local execution, not CI" }, "sources": [ { "title": "A proteomic landscape of pharmacologic perturbations for functional relevance", "url": "https://doi.org/10.1016/j.jpha.2023.08.021", "methods_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC10859532/", "dataset_url": "https://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD043263" }, { "title": "Supplementary tables S1–S3", "urls": [ "https://ars.els-cdn.com/content/image/1-s2.0-S2095177923002101-mmc2.xlsx", "https://ars.els-cdn.com/content/image/1-s2.0-S2095177923002101-mmc3.xlsx", "https://ars.els-cdn.com/content/image/1-s2.0-S2095177923002101-mmc4.xlsx" ], "use": "source attribution only; original workbooks and full derived matrices are not redistributed here" } ], "license_note": "The article is CC BY-NC-ND 4.0. A separate supplementary-data license has not been confirmed. This public file contains only MAEN method descriptions and numerical aggregates.", "limitations": [ "within-study shared-control/batch dependence", "candidate-specific observed feature intersections", "single value permutation is a diagnostic, not a p-value", "external signatures were selected and used by the source authors", "S2 ratio units inferred from methods, not declared in workbook", "no unmeasured-compound or therapeutic efficacy evaluation", "A high missing-mask top-5 score indicates measurement-group information; retrieval does not isolate a causal protein effect.", "No unmeasured compounds, cancer-selective toxicity, human safety or clinical efficacy were evaluated." ] }