{"id":"W2941874147","doi":"10.1007/s40484-019-0166-9","title":"Current challenges and solutions of <i><b>de novo</b></i> assembly","year":2019,"lang":"en","type":"article","venue":"Quantitative Biology","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":96,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"Higher Education Discipline Innovation Project; Fundamental Research Funds for the Central Universities; Project 211; Central South University; National Natural Science Foundation of China","keywords":"De Bruijn graph; Sequence assembly; Computer science; Hybrid genome assembly; DNA sequencing; Graph; Computational biology; Biology; Theoretical computer science; Reference genome; Gene","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":[],"consensus_categories":[],"category_scores_codex":[0.0001550265,0.00008913036,0.0001502172,0.00002861717,0.00002870927,0.00000184158,0.00007612365,0.00007058016,0.000003430187],"category_scores_gemma":[0.00004851466,0.00007967302,0.00004060237,0.00002276234,0.0001344118,4.402313e-7,0.00011202,0.00003993385,0.000007234435],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000003735413,"about_ca_system_score_gemma":0.00003457717,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00000693408,"about_ca_topic_score_gemma":0.00002510189,"domain_scores_codex":[0.999394,0.00007102582,0.000120038,0.0002170589,0.00002271406,0.000175181],"domain_scores_gemma":[0.9996594,0.00004327965,0.00006420763,0.0001345371,0.00006919156,0.00002942447],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.00003402071,0.00003750001,0.009322549,0.00002869516,0.00006365343,8.284757e-8,0.0002278433,0.000004474869,0.8672504,0.118219,0.00005484532,0.00475684],"study_design_scores_gemma":[0.002722619,0.007169624,0.5208198,0.00007873724,0.00009998644,0.00003878297,0.002723233,0.0001317401,0.1596987,0.02458038,0.2809756,0.0009607463],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.906901,0.09091317,0.0004448003,0.0001586285,0.0001662617,0.0000969578,0.00003431807,0.000001639872,0.001283242],"genre_scores_gemma":[0.9838656,0.01450743,0.001496666,0.00003635066,0.00003764225,0.000009126919,0.00001317034,0.000007242682,0.00002679658],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7075517,"threshold_uncertainty_score":0.3248971,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05252731691639779,"score_gpt":0.3162761012553276,"score_spread":0.2637487843389298,"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."}}