FixVAF

FixVAF corrects reference bias introduced by the Illumina Isaac aligner to produce unbiased variant allele frequency (VAF) estimates for cancer genomics analyses.

Key Features:

• Bias correction: Targets reference bias caused by preferential soft clipping of reads supporting non-reference alleles by the Illumina Isaac aligner, producing corrected VAFs for downstream analysis.
• Clipping strategy: Mitigates bias by clipping all reads at variant positions by 5 bases instead of performing computationally intensive realignment.
• Inputs and outputs: Accepts VCF, BAM and FASTA inputs and outputs a modified VCF with adjusted VAFs and allele counts.
• Support for NGS copy number algorithms: Provides allele counts at specified genomic locations to support next-generation sequencing copy number analyses.
• Implementation: Implemented in Python 3 and requires the psam library.

Scientific Applications:

• Cancer cell fraction estimation: Enables more accurate estimation of cancer cell fractions (CCFs) by correcting VAF bias.
• Clonal architecture and mutational timing: Improves inference of clonal architecture and mutational timing in cancer genomes through unbiased VAFs.
• Large-scale sequencing projects: Applicable to large-scale sequencing efforts (e.g., The 100,000 Genomes Project) where VAF accuracy affects downstream analyses.
• Copy number analyses: Supports NGS copy number algorithms by supplying allele counts at genomic loci.

Methodology:

FixVAF takes a VCF file, a BAM file and a FASTA file, clips all reads at variant positions by 5 bases to adjust read alignments and reduce reference bias introduced by the Illumina Isaac aligner, and outputs a modified VCF with adjusted VAFs and allele counts; the implementation is in Python 3 and requires the psam library.