SMuRFv1.4
We have updated the SMuRF models for SNV and Indels using bcbio-v1.0.3. The models are trained on the existing two manually curated Goldsets + the same datasets with adjusted normal coverage and tumor purity. This should improve the overall robustness of SMuRF.
Support for MuTect1.0 has been terminated and we are using the latest MuTect2 according to bcbio-nextgen.
Added the exclusion cases where an individual caller fails to predict any mutations. We still require the input from all 4 callers as each caller provides their features for prediction.
The annotation step for coding regions has been re-written to ensure that SNVs and Indels will be considered as individual cases.