caller_path

[stroke1989]

Hi
Thanks for this excellent tool. caller_path the path of the 3rd-party variant calling program; in demo, Strelka2 was used. could gatk be uesd as caller?

[fanxinping]

You can use another variant calling program in theory, but it's a little difficult. After the SNP calling, het_snp_select step in Accucopy will select the heterozygous SNPs based on the SNP calling results. So, If you use gatk, you should make sure the output of gatk has the same format of Strelka2's output.

The following steps you need to make gatk work in Accucopy:

1. Call the gatk by hand and get the SNPs result.
2. Convert the SNPs result to the same format of Strelka2's output. The header and record format of vcf file must be same, and you can build a program to do it. This step may not be as difficult as you think, because vcf file has their own format standard.
3. Gzip the converted file and rename it to ${strelka_output_dir}/results/variants/variants.vcf.gz. ${strelka_output_dir} is default to strelka_snp , which means the strelka_snp directory under the folder where the tumor file exists if you don't specify the snp_output_dir argument. ${strelka_output_dir} is equal to the snp_output_dir argument.
4. Run the Accucopy but add the argument -s 2, and the Accucopy will skip SNP_calling step and use the ${strelka_output_dir}/results/variants/variants.vcf.gz as the SNPs result directly.

Finally, I recommend use Strelka2, because it's easy to use and we have tested it under the impure condition of tumor tissue.