{"id":"W4396922121","doi":"10.1101/2024.05.13.593988","title":"plantMASST - Community-driven chemotaxonomic digitization of plants","year":2024,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Genetics, Bioinformatics, and Biomedical Research","field":"Biochemistry, Genetics and Molecular Biology","cited_by":27,"is_retracted":false,"has_abstract":true,"ca_institutions":"Wilfrid Laurier University","funders":"","keywords":"Digitization; Botany; Biology; Computer science; Telecommunications","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001338891,0.001022375,0.001205435,0.00681682,0.0008634565,0.002013023,0.001064289,0.0006409385,0.003704241],"category_scores_gemma":[0.003832256,0.0005585236,0.001062516,0.006791758,0.0003485277,0.001276814,0.002739324,0.00116575,0.004134541],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008498997,"about_ca_system_score_gemma":0.001459944,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005422912,"about_ca_topic_score_gemma":0.01322315,"domain_scores_codex":[0.9987778,0.0001158269,0.00009439891,0.0005646673,0.0003332953,0.0001140613],"domain_scores_gemma":[0.9981172,0.0002235055,0.0003057266,0.0006058457,0.0005432729,0.0002044672],"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":[0.001689486,0.0002970019,0.07728328,0.003008872,0.0008330769,0.001156305,0.00165208,0.01239375,0.607826,0.006359032,0.06325829,0.2242427],"study_design_scores_gemma":[0.0002950196,0.0003624158,0.3168775,0.0005689293,0.0005583116,0.002822116,0.001300031,0.08227465,0.1760363,0.02368548,0.3948731,0.0003460387],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"software","genre_scores_codex":[0.3090415,0.003266623,0.1311516,0.0008230663,0.0001966365,0.0003033299,0.5178483,0.02459296,0.01277595],"genre_scores_gemma":[0.1726041,0.001628968,0.1986072,0.000260169,0.00007430871,0.0004009071,0.6224853,0.00173228,0.002206746],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.00681682,"threshold_uncertainty_score":0.01239192,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01979758659384378,"score_gpt":0.2444697552074004,"score_spread":0.2246721686135566,"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."}}