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.

SMuRFv1.3

We have corrected several bugs within the "cdsannotation" function to generate output even if there are no matching cds trancripts found in the snv/indel predictions.

Previously working with MuTect v1 and SMuRF is unable to read the same variables, hence an exception has been created.

General optimization of the code for higher efficiency.

Conflicts of snv-indel mixed calls from different callers:
We have included snv-indel mixed calls to be predicted by both the SNV and indel models, so as not lose any information from the callers. Hence it is possible to find a few indel calls in the snv data file set.

SMuRFv1.2

Even out unnecessary processing to provide a smoother and faster running experience.
Output is generated to give more information - see readme.Rmd

SMuRFv1.1

New models updated with bcbio-v1.0.3 output - models optimized
Bugs removed for:
-pre-processing file writing (removed)
-VarScan "SpvFreq" are considered PASSED calls
-Other minor bug fixes in code

SMuRFv1.0.1

A few bug changes to existing snv and indel models built under bcbio-v0.9.9

SMuRF v1.0

First working version of SMuRF