vcfgl: a flexible genotype likelihood simulator for VCF/BCF files

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Accurate genotype uncertainty quantification is crucial for reliable genetic inferences from next-generation sequencing (NGS) data.
• Genotype Likelihoods (GLs) model uncertainty in base calls but their estimation can be affected by errors and biases.
• The impact of GL estimation biases and model choices on downstream analyses is not fully understood.

Purpose of the Study:

• To introduce vcfgl, a versatile tool for simulating genotype likelihoods with simulated read data.
• To provide a framework for investigating uncertainties and biases in GL quantification.
• To facilitate a deeper understanding of how these factors impact downstream analytical methods.

Main Methods:

• vcfgl simulates genotype likelihoods (GLs) using various established GL models.
• It incorporates simulation of quality score errors using a Beta distribution.
• The tool is compatible with simulators like msprime and SLiM, outputting data in pileup, VCF/BCF, and gVCF formats.

Main Results:

• vcfgl enables simulation and investigation of genotype likelihood uncertainties and biases.
• Simulations demonstrate vcfgl's utility in benchmarking GL-based methods.
• The software supports diverse applications through multiple output formats.

Conclusions:

• vcfgl offers a valuable framework for assessing GL quantification accuracy.
• It aids in understanding and mitigating biases in genetic data analysis.
• The tool enhances the reliability of genetic inferences derived from NGS data.