rewirebio.iobenchmarks
Contested result

LRT and SIFT: Comprehensive assessment of computational algorithms in predicting cancer driver mutations reverses the order

  • Low severity
  • Not yet reviewed
  • Ordering reversal
  • Different tasks across studies

Comprehensive assessment of computational algorithms in predicting cancer driver mutations ranks LRT above SIFT in 16 of 19 comparisons; other studies rank SIFT above LRT in 15 of 17.

An automatic lead, not a finding of error. The results below are shown as published, everywhere on the site, until a reviewer decides otherwise.

Which studies agree

4 papers rank SIFT above LRT in 15 of their 17 comparisons and LRT higher in 2. Comprehensive assessment… ranks LRT above SIFT in 16 of its 20 comparisons and SIFT higher in 3; 1 is a tie.

Papers grouped by which system they rank higher: SIFT or LRTRank SIFT higher: Livesey and Marsh 2025, Brandes et al. 2023, ProteinGym AMFR assay-level Spearman…. Rank LRT higher: Comprehensive assessment… (disagrees). Mixed: Clinical likelihood ratios and balanced….Rank SIFT higherMixedRank LRT higherLivesey and Marsh 2025Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)Brandes et al. 2023Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methodsProteinGym AMFRassay-level Spearman…ProteinGym AMFR assay-level Spearman comparisonClinical likelihoodratios and balanced…Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genesComprehensive assessment…Comprehensive assessment of computational algorithms in predicting cancer driver mutations
  • Rank SIFT higher (3): Livesey and Marsh 2025, Brandes et al. 2023, ProteinGym AMFR assay-level Spearman…
  • Rank LRT higher (1): Comprehensive assessment… (disagrees)
  • Mixed (1): Clinical likelihood ratios and balanced…

Each paper sits where at least two thirds of its comparisons point; Mixed means neither model reaches two thirds. The larger, outlined dot is the paper that disagrees.

Comparisons in which each paper scores SIFT or LRT higherComprehensive assessment… (disagrees): SIFT higher in 3, LRT higher in 16, 1 tie. Livesey and Marsh 2025: SIFT higher in 7, LRT higher in 0, 0 ties. Clinical likelihood ratios and balanced…: SIFT higher in 3, LRT higher in 2, 0 ties. Brandes et al. 2023: SIFT higher in 4, LRT higher in 0, 0 ties. ProteinGym AMFR assay-level Spearman…: SIFT higher in 1, LRT higher in 0, 0 ties.Disagrees: Comprehensive assessment…Comprehensive assessment of computational algorithms in predicting cancer driver mutations3 SIFT higher31 tie16 LRT higher16Livesey and Marsh 2025Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)7 SIFT higher7Clinical likelihood ratios and balanced…Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes3 SIFT higher32 LRT higher2Brandes et al. 2023Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods4 SIFT higher4ProteinGym AMFR assay-level Spearman…ProteinGym AMFR assay-level Spearman comparison1 SIFT higher1
SIFT higherTieLRT higherNumbers are counts of comparisons; the outlined bar is the paper that disagrees.
Papers compared (5)
PaperRoleRanks SIFT higherRanks LRT higherTies
Comprehensive assessment of computational algorithms in predicting cancer driver mutations and 4 more records · DOIDisagrees3161
Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML) and 1 more record · DOIAgrees700
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes and 1 more record · DOIAgrees320
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · DOIAgrees400
ProteinGym AMFR assay-level Spearman comparison and 4 more records · RepositoryAgrees100
Comparisons in Comprehensive assessment… (20)
ProtocolDatasetMetricSIFTLRTHigher
OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Accuracy (OncoKB benchmark (positive set not stated), default categorical calls), fraction0.610.66LRT
OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Negative predictive value (OncoKB benchmark (positive set not stated), default categorical calls), fraction0.750.86LRT
OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Precision (OncoKB benchmark (positive set not stated), default categorical calls), fraction0.580.6LRT
OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Recall (OncoKB benchmark (positive set not stated), default categorical calls), fraction0.890.93LRT
OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Specificity (OncoKB benchmark (positive set not stated), default categorical calls), fraction0.410.38SIFT
OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Accuracy (OncoKB benchmark (positive set not stated), median-score threshold), fraction0.610.65LRT
OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Negative predictive value (OncoKB benchmark (positive set not stated), median-score threshold), fraction0.610.67LRT
OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Precision (OncoKB benchmark (positive set not stated), median-score threshold), fraction0.610.63LRT
OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Recall (OncoKB benchmark (positive set not stated), median-score threshold), fraction0.630.73LRT
OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)OncoKB-annotated somatic missense mutations (oncogenic, likely oncogenic and likely neutral)Specificity (OncoKB benchmark (positive set not stated), median-score threshold), fraction0.60.57SIFT
Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Accuracy (cell viability drivers versus neutral, default categorical calls), fraction0.580.58Tie
Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Negative predictive value (cell viability drivers versus neutral, default categorical calls), fraction0.660.71LRT
Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Precision (cell viability drivers versus neutral, default categorical calls), fraction0.540.56LRT
Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Recall (cell viability drivers versus neutral, default categorical calls), fraction0.80.93LRT
Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Specificity (cell viability drivers versus neutral, default categorical calls), fraction0.480.2SIFT
Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Accuracy (cell viability drivers versus neutral, median-score threshold), fraction0.60.65LRT
Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Negative predictive value (cell viability drivers versus neutral, median-score threshold), fraction0.610.68LRT
Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Precision (cell viability drivers versus neutral, median-score threshold), fraction0.580.63LRT
Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Recall (cell viability drivers versus neutral, median-score threshold), fraction0.670.74LRT
Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Missense mutations with Ba/F3 and MCF10A cell viability calls (published and new)Specificity (cell viability drivers versus neutral, median-score threshold), fraction0.530.56LRT
Comparisons in Livesey and Marsh 2025 (7)
ProtocolDatasetMetricSIFTLRTHigher
Livesey and Marsh 2025 clinical benchmark: pairwise AUROC win rate, ClinVar pathogenic versus gnomAD missenseLivesey and Marsh 2025 clinical set: ClinVar pathogenic versus gnomAD v4.1 missense (819 proteins)Pairwise win rate (average pairwise win rate against all other VEPs, per-protein AUROC on shared variants, ClinVar P/LP versus gnomAD v4.1 missense, proteins with at least 10 of each), percent4918.8SIFT
Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, all 36 DMS setsLivesey and Marsh 2025 human DMS benchmark set (36 proteins)Pairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, 36 human DMS sets), percent52.213.6SIFT
Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, all 36 DMS setsLivesey and Marsh 2025 human DMS benchmark set (36 proteins)Pairwise win rate (average pairwise win rate against the five top-ranked VEPs (column top_5_winrate; the five are not listed in the table), |Spearman| with DMS on shared variants, 36 human DMS sets), percent6.851.22SIFT
Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, direct assays onlyLivesey and Marsh 2025 human DMS benchmark set (36 proteins)Pairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, direct DMS assays only), percent49.915.4SIFT
Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, indirect assays onlyLivesey and Marsh 2025 human DMS benchmark set (36 proteins)Pairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, indirect DMS assays only), percent53.612.3SIFT
Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, SNV-accessible substitutions onlyLivesey and Marsh 2025 human DMS benchmark set (36 proteins)Pairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, SNV-accessible substitutions only), percent52.213.6SIFT
Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, SNV-accessible substitutions with gaps filledLivesey and Marsh 2025 human DMS benchmark set (36 proteins)Pairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS, SNV-accessible substitutions, missing predictions filled with the VEP's most benign score), percent53.614.3SIFT
Comparisons in Clinical likelihood ratios and balanced… (5)
ProtocolDatasetMetricSIFTLRTHigher
BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)BRCA1 functional truth set: 1641 missense variants classed deleterious or tolerated by HAP1 cell survival after saturation genome editingBenignity likelihood ratio (Source 'negative likelihood ratio' (TNR/FNR): likelihood ratio for benignity of a tolerated call, BRCA1 functional truth set), unitless23.92.99SIFT
BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)BRCA1 functional truth set: 1641 missense variants classed deleterious or tolerated by HAP1 cell survival after saturation genome editingFalse negatives (Functionally deleterious BRCA1 variants called tolerated), count385SIFT
BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)BRCA1 functional truth set: 1641 missense variants classed deleterious or tolerated by HAP1 cell survival after saturation genome editingLikelihood ratio (Positive likelihood ratio of a deleterious call (TPR/FPR), BRCA1 functional truth set), unitless1.232.46LRT
BRCA2 functional truth set, binary in silico calls (Cubuk et al. 2021)BRCA2 functional truth set: 188 missense variants classed deleterious or tolerated by Homology-directed repairFalse negatives (Functionally deleterious BRCA2 variants called tolerated), count02SIFT
BRCA2 functional truth set, binary in silico calls (Cubuk et al. 2021)BRCA2 functional truth set: 188 missense variants classed deleterious or tolerated by Homology-directed repairLikelihood ratio (Positive likelihood ratio of a deleterious call (TPR/FPR), BRCA2 functional truth set), unitless1.241.38LRT
Comparisons in Brandes et al. 2023 (4)
ProtocolDatasetMetricSIFTLRTHigher
ClinVar missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)ClinVar missense pathogenic vs benign, primary isoforms (Brandes et al. 2023, from the EVE portal)AUPRC (balanced PRC-AUC: mean over 100 random downsamples to equal pathogenic and benign counts (80% of the minority class); zero-shot), unitless0.8720.779SIFT
ClinVar missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)ClinVar missense pathogenic vs benign, primary isoforms (Brandes et al. 2023, from the EVE portal)AUROC (global ROC-AUC over all variants; zero-shot), unitless0.8740.753SIFT
HGMD/gnomAD missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)HGMD disease-causing vs gnomAD common missense variants, primary isoforms (Brandes et al. 2023)AUPRC (balanced PRC-AUC: mean over 100 random downsamples to equal pathogenic and benign counts (80% of the minority class); zero-shot), unitless0.8820.787SIFT
HGMD/gnomAD missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)HGMD disease-causing vs gnomAD common missense variants, primary isoforms (Brandes et al. 2023)AUROC (global ROC-AUC over all variants; zero-shot), unitless0.8830.76SIFT
Comparisons in ProteinGym AMFR assay-level Spearman… (1)
ProtocolDatasetMetricSIFTLRTHigher
ProteinGym clinical substitutions, zero-shot, average AUC (commit 144fe22)ProteinGym clinical substitution benchmark at commit 144fe22AUROC (ProteinGym clinical substitutions, zero-shot, Average_AUC over proteins), unitless0.8780.805SIFT

What disagrees

Other studies
rank SIFT above LRT in 15 of 17 comparisons
This study
ranks LRT above SIFT in 16 of 19 comparisons
Compared
LRT, SIFT
Study that disagrees
Comprehensive assessment of computational algorithms in predicting cancer driver mutations, Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB), Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability), Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability), Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)
Other studies
10 studiesClinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes, Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx), Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods, ProteinGym AMFR assay-level Spearman comparison, Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML), Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8), ProteinGym clinical supervised substitutions AUC summary at commit 144fe22, ProteinGym clinical zero-shot indels AUC summary at commit 144fe22, ProteinGym clinical zero-shot substitutions AUC summary at commit 144fe22, ProteinGym zero-shot DMS substitutions Spearman summary at commit 144fe22
Rule
Ordering reversal; different tasks across studies
Model identity
Matched across studies partly through links not yet reviewed

The compared numbers

Each value links to its result record and source. Compare values within one protocol; the rule compares rankings, not raw values across protocols.

In the study that disagrees (41)
StudyToolProtocol or benchmarkMetricValue
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Accuracy (OncoKB benchmark (positive set not stated), default categorical calls)0.66
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Negative predictive value (OncoKB benchmark (positive set not stated), default categorical calls)0.86
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Precision (OncoKB benchmark (positive set not stated), default categorical calls)0.6
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Recall (OncoKB benchmark (positive set not stated), default categorical calls)0.93
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Specificity (OncoKB benchmark (positive set not stated), default categorical calls)0.38
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Accuracy (OncoKB benchmark (positive set not stated), median-score threshold)0.65
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Negative predictive value (OncoKB benchmark (positive set not stated), median-score threshold)0.67
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Precision (OncoKB benchmark (positive set not stated), median-score threshold)0.63
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Recall (OncoKB benchmark (positive set not stated), median-score threshold)0.73
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)LRT (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Specificity (OncoKB benchmark (positive set not stated), median-score threshold)0.57
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Negative predictive value (cell viability drivers versus neutral, default categorical calls)0.71
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Precision (cell viability drivers versus neutral, default categorical calls)0.56
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Recall (cell viability drivers versus neutral, default categorical calls)0.93
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Specificity (cell viability drivers versus neutral, default categorical calls)0.2
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Accuracy (cell viability drivers versus neutral, median-score threshold)0.65
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Negative predictive value (cell viability drivers versus neutral, median-score threshold)0.68
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Precision (cell viability drivers versus neutral, median-score threshold)0.63
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Recall (cell viability drivers versus neutral, median-score threshold)0.74
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)LRT (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Specificity (cell viability drivers versus neutral, median-score threshold)0.56
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)SIFT (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Accuracy (OncoKB benchmark (positive set not stated), default categorical calls)0.61
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)SIFT (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Negative predictive value (OncoKB benchmark (positive set not stated), default categorical calls)0.75
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)SIFT (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Recall (OncoKB benchmark (positive set not stated), default categorical calls)0.89
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)SIFT (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Accuracy (OncoKB benchmark (positive set not stated), median-score threshold)0.61
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)SIFT (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Negative predictive value (OncoKB benchmark (positive set not stated), median-score threshold)0.61
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)SIFT (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Recall (OncoKB benchmark (positive set not stated), median-score threshold)0.63
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability)SIFT (Chen et al. 2020)Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Negative predictive value (cell viability drivers versus neutral, default categorical calls)0.66
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability)SIFT (Chen et al. 2020)Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Recall (cell viability drivers versus neutral, default categorical calls)0.8
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)SIFT (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Specificity (cell viability drivers versus neutral, median-score threshold)0.53
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)SIFT4G (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Accuracy (OncoKB benchmark (positive set not stated), default categorical calls)0.61
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)SIFT4G (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Precision (OncoKB benchmark (positive set not stated), default categorical calls)0.58
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 10: Performance metrics of 17 algorithms, default categories, benchmark 2 (OncoKB)SIFT4G (Chen et al. 2020)OncoKB benchmark (positive set not stated), default categorical calls (Chen et al. 2020 Additional file 10)Specificity (OncoKB benchmark (positive set not stated), default categorical calls)0.41
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)SIFT4G (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Accuracy (OncoKB benchmark (positive set not stated), median-score threshold)0.61
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)SIFT4G (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Negative predictive value (OncoKB benchmark (positive set not stated), median-score threshold)0.61
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)SIFT4G (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Precision (OncoKB benchmark (positive set not stated), median-score threshold)0.61
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 9: Performance metrics of 33 algorithms, median-score threshold, benchmark 2 (OncoKB)SIFT4G (Chen et al. 2020)OncoKB benchmark (positive set not stated), median-score threshold (Chen et al. 2020 Additional file 9)Specificity (OncoKB benchmark (positive set not stated), median-score threshold)0.6
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability)SIFT4G (Chen et al. 2020)Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Precision (cell viability drivers versus neutral, default categorical calls)0.54
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 22: Performance metrics of 17 algorithms, default categories, benchmark 5 (cell viability)SIFT4G (Chen et al. 2020)Cell viability drivers versus neutral, default categorical calls (Chen et al. 2020 Additional file 22)Specificity (cell viability drivers versus neutral, default categorical calls)0.48
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)SIFT4G (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Accuracy (cell viability drivers versus neutral, median-score threshold)0.6
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)SIFT4G (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Negative predictive value (cell viability drivers versus neutral, median-score threshold)0.61
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)SIFT4G (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Precision (cell viability drivers versus neutral, median-score threshold)0.58
Comprehensive assessment of computational algorithms in predicting cancer driver mutations; Chen et al. 2020, Additional file 21: Performance metrics of 33 algorithms, median-score threshold, benchmark 5 (cell viability)SIFT4G (Chen et al. 2020)Cell viability drivers versus neutral, median-score threshold (Chen et al. 2020 Additional file 21)Recall (cell viability drivers versus neutral, median-score threshold)0.67
In the other studies (34)
StudyToolProtocol or benchmarkMetricValue
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)LRT (Cubuk et al. 2021)BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)Benignity likelihood ratio (Source 'negative likelihood ratio' (TNR/FNR): likelihood ratio for benignity of a tolerated call, BRCA1 functional truth set)2.99
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)LRT (Cubuk et al. 2021)BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)False negatives (Functionally deleterious BRCA1 variants called tolerated)85
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)LRT (Cubuk et al. 2021)BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)Likelihood ratio (Positive likelihood ratio of a deleterious call (TPR/FPR), BRCA1 functional truth set)2.46
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)LRT (Cubuk et al. 2021)BRCA2 functional truth set, binary in silico calls (Cubuk et al. 2021)False negatives (Functionally deleterious BRCA2 variants called tolerated)2
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)LRT (Cubuk et al. 2021)BRCA2 functional truth set, binary in silico calls (Cubuk et al. 2021)Likelihood ratio (Positive likelihood ratio of a deleterious call (TPR/FPR), BRCA2 functional truth set)1.38
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)SIFT (Cubuk et al. 2021)BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)Benignity likelihood ratio (Source 'negative likelihood ratio' (TNR/FNR): likelihood ratio for benignity of a tolerated call, BRCA1 functional truth set)23.9
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)SIFT (Cubuk et al. 2021)BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)False negatives (Functionally deleterious BRCA1 variants called tolerated)3
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)SIFT (Cubuk et al. 2021)BRCA1 functional truth set, binary in silico calls (Cubuk et al. 2021)Likelihood ratio (Positive likelihood ratio of a deleterious call (TPR/FPR), BRCA1 functional truth set)1.23
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)SIFT (Cubuk et al. 2021)BRCA2 functional truth set, binary in silico calls (Cubuk et al. 2021)False negatives (Functionally deleterious BRCA2 variants called tolerated)0
Clinical likelihood ratios and balanced accuracy for 44 in silico tools against multiple large-scale functional assays of cancer susceptibility genes; Cubuk et al. 2021, Supplementary tables 1-13 (41436_2021_1265_MOESM3_ESM.xlsx)SIFT (Cubuk et al. 2021)BRCA2 functional truth set, binary in silico calls (Cubuk et al. 2021)Likelihood ratio (Positive likelihood ratio of a deleterious call (TPR/FPR), BRCA2 functional truth set)1.24
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods; Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023)LRT (Brandes et al. 2023 clinical benchmarks)ClinVar missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)AUPRC (balanced PRC-AUC: mean over 100 random downsamples to equal pathogenic and benign counts (80% of the minority class); zero-shot)0.779
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods; Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023)LRT (Brandes et al. 2023 clinical benchmarks)ClinVar missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)AUROC (global ROC-AUC over all variants; zero-shot)0.753
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods; Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023)SIFT (Brandes et al. 2023 clinical benchmarks)ClinVar missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)AUPRC (balanced PRC-AUC: mean over 100 random downsamples to equal pathogenic and benign counts (80% of the minority class); zero-shot)0.872
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods; Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023)SIFT (Brandes et al. 2023 clinical benchmarks)ClinVar missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)AUROC (global ROC-AUC over all variants; zero-shot)0.874
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods; Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023)LRT (Brandes et al. 2023 clinical benchmarks)HGMD/gnomAD missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)AUPRC (balanced PRC-AUC: mean over 100 random downsamples to equal pathogenic and benign counts (80% of the minority class); zero-shot)0.787
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods; Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023)LRT (Brandes et al. 2023 clinical benchmarks)HGMD/gnomAD missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)AUROC (global ROC-AUC over all variants; zero-shot)0.76
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods; Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023)SIFT (Brandes et al. 2023 clinical benchmarks)HGMD/gnomAD missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)AUPRC (balanced PRC-AUC: mean over 100 random downsamples to equal pathogenic and benign counts (80% of the minority class); zero-shot)0.882
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods; Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023)SIFT (Brandes et al. 2023 clinical benchmarks)HGMD/gnomAD missense pathogenicity, variants scored by all 19 eligible methods (Brandes et al. 2023 Fig. 2e-f, Supplementary Table 2)AUROC (global ROC-AUC over all variants; zero-shot)0.883
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)LRT (Livesey and Marsh 2025)Livesey and Marsh 2025 clinical benchmark: pairwise AUROC win rate, ClinVar pathogenic versus gnomAD missensePairwise win rate (average pairwise win rate against all other VEPs, per-protein AUROC on shared variants, ClinVar P/LP versus gnomAD v4.1 missense, proteins with at least 10 of each)18.8
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)SIFT4G (Livesey and Marsh 2025)Livesey and Marsh 2025 clinical benchmark: pairwise AUROC win rate, ClinVar pathogenic versus gnomAD missensePairwise win rate (average pairwise win rate against all other VEPs, per-protein AUROC on shared variants, ClinVar P/LP versus gnomAD v4.1 missense, proteins with at least 10 of each)49
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)LRT (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, all 36 DMS setsPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, 36 human DMS sets)13.6
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)LRT (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, all 36 DMS setsPairwise win rate (average pairwise win rate against the five top-ranked VEPs (column top_5_winrate; the five are not listed in the table), |Spearman| with DMS on shared variants, 36 human DMS sets)1.22
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)SIFT4G (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, all 36 DMS setsPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, 36 human DMS sets)52.2
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)SIFT4G (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, all 36 DMS setsPairwise win rate (average pairwise win rate against the five top-ranked VEPs (column top_5_winrate; the five are not listed in the table), |Spearman| with DMS on shared variants, 36 human DMS sets)6.85
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)LRT (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, direct assays onlyPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, direct DMS assays only)15.4
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)SIFT4G (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, direct assays onlyPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, direct DMS assays only)49.9
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)LRT (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, indirect assays onlyPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, indirect DMS assays only)12.3
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)SIFT4G (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, indirect assays onlyPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, indirect DMS assays only)53.6
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)LRT (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, SNV-accessible substitutions with gaps filledPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS, SNV-accessible substitutions, missing predictions filled with the VEP's most benign score)14.3
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)SIFT4G (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, SNV-accessible substitutions with gaps filledPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS, SNV-accessible substitutions, missing predictions filled with the VEP's most benign score)53.6
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)LRT (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, SNV-accessible substitutions onlyPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, SNV-accessible substitutions only)13.6
Livesey and Marsh 2025, Supplementary Material 2 (Tables S1-S8); Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML)SIFT4G (Livesey and Marsh 2025)Livesey and Marsh 2025 DMS benchmark: pairwise |Spearman| win rate, SNV-accessible substitutions onlyPairwise win rate (average pairwise win rate against all other VEPs, |Spearman| with DMS on shared variants, SNV-accessible substitutions only)52.2
ProteinGym clinical zero-shot substitutions AUC summary at commit 144fe22LRT [ProteinGym 144fe22 clinical summary]ProteinGym clinical substitutions, zero-shot, average AUC (commit 144fe22)AUROC (ProteinGym clinical substitutions, zero-shot, Average_AUC over proteins)0.805
ProteinGym clinical zero-shot substitutions AUC summary at commit 144fe22SIFT [ProteinGym 144fe22 clinical summary]ProteinGym clinical substitutions, zero-shot, average AUC (commit 144fe22)AUROC (ProteinGym clinical substitutions, zero-shot, Average_AUC over proteins)0.878

Follow-up

Current state: Not yet reviewed.

No one has reviewed this finding yet.

How to help resolve this

If you know why these studies disagree, for example a different population, metric or preprocessing step, or you have a corrected value, tell us through the contribution form. A reviewer checks every submission against its source before anything changes, and the step is added to the follow-up above.

Finding discrepancy-96662aae7a60911f · All contested results · How findings are made and followed up