Hi

I am trying to identify somatic single nucleotide variants on FFPE samples and have tried various tools to maximize the overlap of variants called in the Fresh and FFPE samples.

LoFreq had the best results (~70% of overlap) where as SomVarIUS resulted in (~10% overlap)

To understand what makes LoFreq significantly better with FFPE samples, I tried to compare the algorithms described in their papers (LoFreq: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3526318/#gks918-B11 SomVarIUS: https://academic.oup.com/bioinformatics/article/32/6/808/1744570) but I am not sure what the difference is since they both follow similar procedures (statistics) to model sequencing errors in order eliminate false positives.

Any help in helping distinguish the two is appreciated!

Thanks