How-to: call variants from RNA-seq data
https://software.broadinstitute.org/gatk/best-practices/workflow?id=11164
Install GATK4
Download the precompiled jar binaries. Set paths.
Adding read groups to bam file
java -jar picard AddOrReplaceReadGroups \
I=sample.filt.bam \
O=sample.rg.bam \
SO=coordinate RGID=id RGLB=library RGPL=platform RGPU=machine RGSM=sample
Marking duplicates
java -jar picard MarkDuplicates \
I=sample.rg.bam \
O=sample.dedupped.bam \
CREATE_INDEX=true VALIDATION_STRINGENCY=SILENT M=output.metrics
Splitting and trimming
gatk SplitNCigarReads \
-R genome.fa \
-I sample.dedupped.bam \
-O sample.split.filtered.bam
Haplotype calling
gatk HaplotypeCaller \
-R genome.fa \
-L 1 2 3 4 5 6 X \
-I sample.split.filtered.bam \
-O sample.filtered.vcf \
--dont-use-soft-clipped-bases \
--standard-min-confidence-threshold-for-calling 20.0