GHGA QC tools

Overview

The two main scripts can be used to compare two BAM (bamqc.sh) or two VCF (vcfqc.sh) files on different metrics. Results of each comparison are collected in a table that has to be specified in the input parameters (collect_table.tsv).

Requirements are:

• Samtools >= 1.6
• Bcftools >= 1.6
• python >= 3
• pandas >= 1.0.0

In addition theres is also a quick wrapper script for comparing two bamfiles (quick_bam_comparison.sh). It first checks md5sums of both files, if they are unequal it runs the picard CompareSAMs tool with stringent settings. These results will not not be collected in a central table. The script will only produce the picard results table.

Requirements are:

• picard >= 2.19.1

Run tools

bamqc.sh bamfile1 bamfile2 collect_bam_table.tsv

vcfqc.sh vcffile1 vcffile2 collect_vcf_table.tsv

note: vcf inputs for vcfqc.sh have to be bgzipped & indexed

also note: you can concatenate multiple comparisons for each bamqc & vcfqc into a single collect_table.tsv.

quick_bam_comparison.sh bamfile1 bamfile2 picard_memory picard_output_file

note: picard_memory has to be in java conform format (see https://docs.oracle.com/javase/7/docs/technotes/tools/solaris/java.html)

Comparison metrics of bamqc.sh and vcfqc.sh

BAM comparison metrics:

• Difference of QC-passed reads (qc_passed_diff)
• Difference of mapped reads (mapped_reads_diff)
• Difference of paired reads (paired_reads_diff)
• Difference of average depth per chromosome (chr1_dp_diff, chr2_dp_diff ... chrY_dp_diff)

VCF comparison metrics:

• Difference of variant counts (vars_diff)
• Difference of SNP counts (snps_diff)
• Differences of indel counts (indels_diff)
• Number of genotype mismatches (gt_mismatches)
• Difference of transition/transversion ratios (Ti/Tv_diff)
• Mean and standard deviation of DP, MQ and QD differences (DP_mean, DP_std, MQ_mean ... QD_std)
• Difference of variant per chromosome (chr1_diff, chr2_diff ... chrY_diff)