{"id":"W4407114045","doi":"10.1186/s12859-025-06050-2","title":"BioLake: an RNA expression analysis framework for prostate cancer biomarker powered by data lakehouse","year":2025,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Prostate Cancer Canada; University of Calgary","funders":"","keywords":"Upload; Computer science; Raw data; Scalability; Flexibility (engineering); Data science; Data pre-processing; Pipeline (software); Data mining; Workflow; Data warehouse; Database; World Wide Web","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002594873,0.000195144,0.0002013512,0.00015343,0.0001546223,0.00009994207,0.0006184662,0.0002593628,0.0000311638],"category_scores_gemma":[0.0001407799,0.0001638398,0.0001035058,0.0005065009,0.00005714264,0.00003194755,0.0002383117,0.00007656501,0.0000033875],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000209064,"about_ca_system_score_gemma":0.0002039774,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001396131,"about_ca_topic_score_gemma":0.00008648564,"domain_scores_codex":[0.9987398,0.0000436515,0.0004259541,0.0003828363,0.0001472835,0.0002604242],"domain_scores_gemma":[0.9980821,0.00002478689,0.0002337301,0.001426848,0.0001269801,0.0001055726],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0008239068,0.0003059156,0.01012975,0.0004266254,0.0007345519,1.589346e-7,0.0003777568,0.0004720929,0.5798833,0.000396326,0.3502804,0.05616929],"study_design_scores_gemma":[0.001725043,0.0001642428,0.003093042,0.0001661607,0.0005877129,5.23992e-7,0.0009732742,0.1457085,0.3026451,0.0004974161,0.5437307,0.0007082899],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0263741,0.001725782,0.9680182,0.0002836941,0.0003947967,0.0006871753,0.001888832,0.00005636984,0.0005710098],"genre_scores_gemma":[0.5317146,0.002932469,0.4281262,0.002687919,0.0003290868,0.0007432561,0.02754691,0.0000906025,0.005828998],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5398921,"threshold_uncertainty_score":0.6681195,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04424208771644974,"score_gpt":0.352946851264794,"score_spread":0.3087047635483443,"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."}}