Code for analysing Nodular Melanoma samples from multi-regional NGS data
Perl script to parse information from a mutect2 vcf files (generated with Gatk 4.3.0) and create a custom .tsv file that is compatible with ANNOVAR execution for annotation and an extra column with summary stats from variant calling. e.g., Is it a PASS or filtered mutation, the DP, AF and AD values, usefull for further scrutiny. Note: For ANNOVAR, whenever there is a deletion or instertion, it is represented as "-". For example, say there is an insertion. In mutect2 vcf file this would be shown us :
| REF | ALT |
|---|---|
| G | GACTACT |
However for ANNOVAR REF would be indicated as "-" and ALT as "ACTACT". Therefore start and end positions of the variant slightly change. The code also accounts for that.
Parse information from samtools/bcftools mpileup. The format looks like: #CHROM POS ID REF ALT QUAL FILTER INFO FORMAT TUMOUR CHROM, POS, REF, ALT, DP4 from INFO are extracted into a tabular delimited .txt file Script takes input and output as arguments
Use the DP4 column which holds integer a,b,c,d ( where a,b are fw/rev reference reads and c,d are fw/rev alt reads) to keep values with at least 10 reads (a+b+c+d), at least 4 tumour reads ( c+d) and at least 10% MAF (c+d+/(a+b+c+d)
Add an extra column to the custom mpileup called MAF, generated from DP4 values
Use Copy Number analysis files to extract copy number per segments and assign local copy numbers to the mpileup file