{"id":"W4392356342","doi":"10.1038/s41587-023-02097-9","title":"Characterizing the impacts of dataset imbalance on single-cell data integration","year":2024,"lang":"en","type":"article","venue":"Nature Biotechnology","topic":"Single-cell and spatial transcriptomics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":45,"is_retracted":false,"has_abstract":false,"ca_institutions":"Ontario Institute for Cancer Research; Lunenfeld-Tanenbaum Research Institute; Simon Fraser University; Vector Institute; University of Toronto; University Health Network","funders":"Natural Sciences and Engineering Research Council of Canada; Government of Canada","keywords":"Data integration; Cluster analysis; Computer science; Inference; Benchmarking; Data mining; Robustness (evolution); Annotation; Computational biology; Biology; Artificial intelligence; Gene; Genetics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.01344542,0.001187619,0.001382048,0.002070115,0.001471964,0.003444318,0.001496874,0.001849328,0.0008178101],"category_scores_gemma":[0.06674837,0.0005844897,0.001378439,0.004664941,0.001297118,0.003655572,0.002675445,0.001862196,0.0005628646],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001633975,"about_ca_system_score_gemma":0.002482294,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005702968,"about_ca_topic_score_gemma":0.006825703,"domain_scores_codex":[0.9878139,0.003058412,0.00123286,0.003573628,0.003380324,0.0009407617],"domain_scores_gemma":[0.9566543,0.02996196,0.002556232,0.005472242,0.004610121,0.000745081],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.006583764,0.001714873,0.3709777,0.001415777,0.00257379,0.001693995,0.000824239,0.2908875,0.1082213,0.006909174,0.02305065,0.1851472],"study_design_scores_gemma":[0.0002269126,0.0007411356,0.1171775,0.0001233548,0.0007883428,0.001424523,0.001205561,0.776273,0.07381758,0.01798479,0.01010896,0.0001283092],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9021074,0.003049154,0.07062336,0.002515473,0.0006710701,0.0002574341,0.01452991,0.003610479,0.002635826],"genre_scores_gemma":[0.9143569,0.0005097203,0.04438915,0.0008621565,0.0002011016,0.0002278293,0.03822831,0.0004954344,0.0007293838],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9865546,"threshold_uncertainty_score":0.07110703,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01714862689806865,"score_gpt":0.2666866661535022,"score_spread":0.2495380392554336,"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."}}