{"id":"W4416980514","doi":"10.26434/chemrxiv-2025-tm1ms","title":"The need for standardisation and improved open (meta)data practices in metaproteomics","year":2025,"lang":"","type":"article","venue":"ChemRxiv","topic":"Advanced Proteomics Techniques and Applications","field":"Chemistry","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Metaproteomics; Interoperability; Workflow; Metadata; Standardization; Data curation; Automation; Limiting","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":["metaresearch","open_science"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.6191025,0.002892043,0.003111346,0.0159043,0.006307499,0.02709889,0.01648602,0.00941753,0.002171298],"category_scores_gemma":[0.6094972,0.002922388,0.005162392,0.01421059,0.01569227,0.05202637,0.02906592,0.02816316,0.00264286],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.01150129,"about_ca_system_score_gemma":0.06778642,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008694666,"about_ca_topic_score_gemma":0.006949815,"domain_scores_codex":[0.5309662,0.2556003,0.1010339,0.02054138,0.08542635,0.006431909],"domain_scores_gemma":[0.1419285,0.310957,0.06668154,0.3034129,0.1647188,0.01230132],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0007717733,0.001475188,0.06074802,0.02301605,0.001745862,0.0009587714,0.02600466,0.007213477,0.03043432,0.2239829,0.09985846,0.5237907],"study_design_scores_gemma":[0.0002528227,0.0007406744,0.02120576,0.03491483,0.0008322386,0.001141642,0.01249538,0.007794406,0.02895161,0.3425424,0.5480174,0.001110837],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02068062,0.01331115,0.7563518,0.1768833,0.005496387,0.005403196,0.003709747,0.006796757,0.0113671],"genre_scores_gemma":[0.03066503,0.004275818,0.9384508,0.0140968,0.0009937006,0.003963579,0.005126438,0.001505769,0.00092206],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.983514,"threshold_uncertainty_score":0.4697144,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08487166913669698,"score_gpt":0.3999366695188863,"score_spread":0.3150650003821893,"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."}}