{"id":"W3121726367","doi":"10.1101/2021.01.14.426572","title":"Data proliferation, reconciliation, and synthesis in viral ecology","year":2021,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Zoonotic diseases and public health","field":"Medicine","cited_by":28,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; University of Saskatchewan; University of Toronto","funders":"Institut de Valorisation des Données; National Science Foundation","keywords":"Human virome; Metadata; Ecology; Rubric; Data science; Microbial ecology; Evolutionary ecology; Biology; Host (biology); Metagenomics; Computer science; World Wide Web; Sociology","routes":{"ca_aff":true,"ca_fund":true,"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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.00122874,0.0003289065,0.0007233255,0.0002614137,0.00008336266,0.0001805443,0.00030918,0.0004894771,0.0002037545],"category_scores_gemma":[0.002702414,0.0003498202,0.00005366045,0.000300738,0.0000809117,0.0002363958,0.0005812067,0.000602878,0.00001424961],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002878628,"about_ca_system_score_gemma":0.004059862,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000128608,"about_ca_topic_score_gemma":0.00005786587,"domain_scores_codex":[0.9972291,0.0002436091,0.0006772851,0.001132361,0.0002775781,0.0004400772],"domain_scores_gemma":[0.9969128,0.0001617661,0.0002539478,0.001780846,0.0004603655,0.0004302779],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0003090105,0.002224153,0.9673058,0.007365135,0.0007101197,0.001029552,0.00008828724,0.00004204758,0.01319293,0.00296985,0.004593584,0.0001695627],"study_design_scores_gemma":[0.001109988,0.00007047441,0.9855056,0.0008130104,0.0002684435,1.561187e-7,0.00002112892,0.003628264,0.001949933,0.000004318005,0.006121672,0.0005070303],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9922767,0.001264963,0.0001858847,0.003985727,0.0006413147,0.000961798,0.0005201256,0.0001270525,0.00003646763],"genre_scores_gemma":[0.9935043,0.0006146639,0.003908569,0.001164473,0.0005097323,0.0002215265,0.000008667042,0.00006208096,0.000006026546],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01819981,"threshold_uncertainty_score":0.9998954,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03099408012562779,"score_gpt":0.2697257310317566,"score_spread":0.2387316509061288,"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."}}