Hi,
I was trying to modify sv-pipeline to be able to classify variant calls of a smaller cohort. calc_params would fail if deletions are not present in the naive Bayes training call set, contradicting the post-merge WDL script's logic. Edits are made so this could run classifications of svtypes other than deletions without crashing due to an empty deletion dataframe. Please review. Thanks
Bohan
Hi, I was trying to modify sv-pipeline to be able to classify variant calls of a smaller cohort. calc_params would fail if deletions are not present in the naive Bayes training call set, contradicting the post-merge WDL script's logic. Edits are made so this could run classifications of svtypes other than deletions without crashing due to an empty deletion dataframe. Please review. Thanks Bohan