Novotyper

This repository contains the code used during the development of a proof of concept for a structural variant genotyper tentatively called Novotyper.

• It was created to investigate whether the ratio of the mapping quality at an alternate contig constructed from a predicted SV to the mapping quality at the reference location (as calculated by Novoalign) could be used to accurately genotype structural variants from short-read WGS data.
• That is, we wanted to see whether the MAPQ ratio could be used to classify the sample as homozygous reference, heterozygous, or homozygous alternate for a given structural variant.

Quick start

1. Obtain a VCF of the structural variants you would like to genotype.

2. Run the Snakemake workflow found in workflow/Snakefile.

   • Modify the config/config.yaml file to point to the correct locations for each input file and executable.

   • The workflow should perform the following steps (but you could run them yourself):

     Tool	Purpose
     novoSV vcf2altscaffold	Use the SVs from the VCF file along with the reference .nix file to generate a FASTA file containing the alt scaffolds.
     novoindex	Index the reference fasta + alt scaffolds fasta to create a new spiked .nix file.
     novoalign	Produce the BAM file by aligning the reads using the spiked .nix file.
     novoSV zabedgraph	Use the alt scaffolds FASTA and the BAM file to obtain the MAPQ bedgraph file.
     novotyper	Use the SV VCF, alt scaffolds FASTA, and MAPQ bedgraph to predict the genotypes and perform some benchmarking.
     novoutil bgzf [OPTIONAL]	Combine reference fasta + alt scaffolds fasta into a new fasta file, bgzip it, and index it with samtools faidx. Do this if you would like to have the spiked reference file to visualize the alignments in IGV.

   • See DAG.

Report

• For a basic description of the reasoning behind each step and the results, please download notebooks/report.html and view it in your browser.

Benchmarks

• HG002 (reads and truth set obtained from GIAB)
• HG005 (reads and truth set obtained from this paper)
• HG00513 (reads and truth set obtained from HGSVC2)

Acknowledgements

This work was part of my internship at Novocraft (July 2024 - September 2024).