Improved algorithms for optical genome mapping workflows in constitutional disease and oncology applications
Laboratory Genetics and Genomics
-
Primary Categories:
- Genomic Medicine
-
Secondary Categories:
- Genomic Medicine
Introduction:
Methods:
Results:
Conclusion:
Introduction
Optical genome mapping (OGM) enables the detection of genomic structural and copy number variants, many of which are missed by next generation sequencing (NGS) technologies and/or cytogenetic techniques. Bionano has developed bioinformatics pipelines to call structural variants (SVs) and copy number variants (CNVs) including the Bionano Solve de novo assembly pipeline for constitutional analysis and the rare variant pipeline for low allele-fraction oncology and other heterogeneous cell mixture applications. In addition, there is a coverage-based copy number component of the pipeline for OGM analysis called FractCNV to find CNVs based on read depth coverage. While these pipelines have been applied in numerous rare inherited disease and cancer studies, there are some limitations with each, therefore, some labs use both SV-calling pipelines in combination to generate the most comprehensive call list.
Methods:
Methods Guided assembly (GA) is a new pipeline that aims to combine the low-allele fraction detection capability of the rare variant pipeline with the more contiguous, whole genome coverage enabled by de novo assembly pipeline. GA uses the reference genome as the initial seed followed by extension, refinement and SV calling steps. Additionally, a new CNV calling tool, SNP-FASST3, improves CNV calling sensitivity down to a level of 10% aberrant cell fraction or 5% variant allele fraction (VAF). The new pipelines have been evaluated through comparison to standard benchmarking datasets and cases with known disease-associated variants to measure their performance.
Results:
Results
Variants that were previously challenging to detect with one or the other previous pipelines are identified efficiently with the new pipelines. Some examples of variants with improved sensitivity include: a low-allele fraction 17p arm loss impacting TP53, chr11 CCND1 rearrangement in lymphoma, CRLF2 deletion and translocation in acute lymphocytic leukemia and t(4:14) IGH translocation seen in multiple myeloma.
Conclusion:
Conclusions
The GA and SNP-FAST3 pipelines offer accurate and sensitive variant-calling, and comparison to existing methods has shown improved analytical performance in sensitivity for variant detection with low VAF and complex variants. New workflow recommendations are also provided for constitutional germline, constitutional mosaic, and oncology applications. For constitutional germline application, one can run the de novo or constitutional guided assembly on 100X raw coverage data to detect variants ≥1kbp. For finding mosaic calls, one would run the low allele-fraction guided assembly on 200X raw coverage data to detect variants ≥3kbp at ≥10% VAF. Finally, for oncology applications, users can run the low allele-fraction guided assembly on 400X raw coverage data to call variants ≥3kbp at ≥5% VAF.