{"id":"W4280519287","doi":"10.12688/f1000research.109080.2","title":"Recommendations for the formatting of Variant Call Format (VCF) files to make plant genotyping data FAIR","year":2022,"lang":"en","type":"preprint","venue":"F1000Research","topic":"Research Data Management Practices","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ontario Institute for Cancer Research","funders":"Biotechnology and Biological Sciences Research Council; Horizon 2020 Framework Programme; German Network for Bioinformatics Infrastructure; Bundesministerium für Bildung und Forschung; European Commission","keywords":"Disk formatting; Metadata; Computer science; Information retrieval; Identifier; World Wide Web; Programming language","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":["scholarly_communication","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.01518688,0.0002942648,0.0003811796,0.0008735666,0.001110148,0.003823592,0.02752641,0.0001009444,0.0002150659],"category_scores_gemma":[0.004200166,0.0002284235,0.00012298,0.0009947784,0.00008932494,0.007291331,0.09889464,0.001514352,0.00003532473],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002426328,"about_ca_system_score_gemma":0.0007841168,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001412479,"about_ca_topic_score_gemma":0.0003831825,"domain_scores_codex":[0.9939482,0.0007252265,0.0008527014,0.001221719,0.002163224,0.001088956],"domain_scores_gemma":[0.9860283,0.005537031,0.0004469139,0.007343462,0.0004120788,0.0002322092],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001919744,0.0002734374,0.00008014405,0.002034269,0.0006961421,0.00005583865,0.003363247,0.005529722,0.0001310525,0.1414138,0.6756019,0.1706284],"study_design_scores_gemma":[0.0001823164,0.00009031266,0.0001832079,0.0001244826,0.00002196526,0.000009127645,0.0007307969,0.322079,0.00007337663,0.001125583,0.675165,0.0002148178],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.00008696219,0.0004516313,0.9310082,0.05019829,0.0007694702,0.004371801,0.00949261,0.0001402379,0.003480769],"genre_scores_gemma":[0.01931477,0.01282211,0.9139546,0.002672402,0.001034748,0.009138975,0.02757432,0.000251607,0.01323649],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.3165493,"threshold_uncertainty_score":0.9972106,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4089725583043259,"score_gpt":0.4605710201384414,"score_spread":0.05159846183411543,"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."}}