Yum, tasty mutations...

MutationTaster2021 Supplementary Data

Random Forest models

Model Features Scores Grid search Precision-recall Random Forest model Sample tree
without_aae .txt .txt .tsv .png .pkl .json .svg
simple_aae .txt .txt .tsv .png .pkl .json .svg
complex_aae .txt .txt .tsv .png .pkl .json .svg
3utr .txt .txt .tsv .png .pkl .json .svg
5utr .txt .txt .tsv .png .pkl .json .svg
All models .tar.gz .tar.gz .tar.gz .tar.gz .tar.gz .tar.gz .tar.gz

Training data

For gnomAD and ClinVar, training and test sets have the format:

chr   pos   ref   alt   transcript

HGMD sets have the format:

HGMD ID   transcript

Training and test sets were generated at transcript level. Thus, a variant can appear in the training and test sets if it affects more than one transcript.

Training set

Model Benign Deleterious
gnomAD ClinVar HGMD
without_aae .tsv.gz .tsv.gz .tsv.gz
simple_aae .tsv.gz .tsv.gz .tsv.gz
complex_aae .tsv.gz .tsv.gz .tsv.gz
3utr .tsv.gz .tsv.gz .tsv.gz
5utr .tsv.gz .tsv.gz .tsv.gz
All models .tar.gz .tar.gz .tar.gz

Test set

Model Benign Deleterious
gnomAD ClinVar HGMD
without_aae .tsv.gz .tsv.gz .tsv.gz
simple_aae .tsv.gz .tsv.gz .tsv.gz
complex_aae .tsv.gz .tsv.gz .tsv.gz
3utr .tsv.gz .tsv.gz .tsv.gz
5utr .tsv.gz .tsv.gz .tsv.gz
All models .tar.gz .tar.gz .tar.gz

Updated

24 March 2021