{"id":"W4366991708","doi":"10.1101/2023.04.21.537695","title":"hipFG: High-throughput harmonization and integration pipeline for functional genomics data","year":2023,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Weston Brain Institute; National Institute on Aging; Alzheimer's Association","keywords":"Workflow; Computer science; Pipeline (software); Normalization (sociology); Harmonization; Data integration; Genomics; Computational biology; Scalability; Genome; Data mining; Database; Biology; Programming language; Genetics; Gene","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.008786905,0.003006608,0.002024339,0.00477433,0.001872702,0.004120138,0.004141155,0.001125339,0.02626855],"category_scores_gemma":[0.01123311,0.002039954,0.002856193,0.004407918,0.001161422,0.002968672,0.006862527,0.003633994,0.02251964],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001596344,"about_ca_system_score_gemma":0.004809097,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00683145,"about_ca_topic_score_gemma":0.006570151,"domain_scores_codex":[0.995481,0.0005909585,0.0004222979,0.001459319,0.00165271,0.0003935769],"domain_scores_gemma":[0.9948866,0.001447904,0.0003492861,0.001867187,0.00103785,0.0004112359],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001434149,0.000160896,0.00633942,0.001230509,0.0005185987,0.0006133847,0.0007078258,0.007090651,0.03903932,0.00724553,0.8047508,0.130869],"study_design_scores_gemma":[0.0008353,0.0002367018,0.02583439,0.0004704852,0.0003962804,0.001193436,0.0004730271,0.08983222,0.1516002,0.05052301,0.6777827,0.0008222443],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"software","genre_gemma":"software","genre_scores_codex":[0.007897287,0.0004641942,0.3323005,0.0009102311,0.0003868208,0.0006059952,0.1537426,0.4989578,0.004734491],"genre_scores_gemma":[0.04508317,0.0004452775,0.4374792,0.001126669,0.0002287176,0.00248382,0.4335768,0.0738331,0.005743208],"genre_candidate":"software","genre_consensus":"software","teacher_disagreement_score":0.02626855,"threshold_uncertainty_score":0.08787704,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04998630637713508,"score_gpt":0.2612267926738121,"score_spread":0.211240486296677,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}