We’ve just released version 2.1.3 (it’s been a while…). This version
brings a few minor changes: Python 3 support, a –force-overwrite
option to call and a best practices snakemake workflow.
This is likely the last version 2 release. We’ve started a complete rewrite which will make the tool much faster, dynamically compiles against htslib etc.
We recently got asked a lot why LoFreq’s VCF output has no FORMAT and SAMPLE columns. The reason for their absence is that they represent genotyping information and current LoFreq versions don’t call genotypes. This shouldn’t stop you from using LoFreq, as genotype information is often not even needed (depending on your analysis). Some downstream tools might required it (e.g. vcf_melt), even though these columns are optional according to the VCF specification (see e.g. section 1.3).
As a workaround, you can just add fake columns in cases where you know that the information is actually not required and for somatic samples you can use (lofreq2_add_sample.py , which comes with LoFreq (note the pysam dependency). The next versions of LoFreq (2.2) will be able to call genotypes.
Here’s the LoFreq poster presented at the EMBL Microbiome conference 2015.
Andreas