Seven Bridges, the industry-leading bioinformatics ecosystem provider, today announced the launch of the GRAF™ Population Solution, a set of workflows, services and population-specific graph references to enable large-scale human population studies through more accurate read alignment and variant discovery. Included in today’s release are an updated population-specific NGS analysis pipeline, population-specific graph construction services and graph references for five major global populations — African, East Asian, South Asian, Admixed American and European. These offerings will augment the GRAF Germline Variant Detection Workflow and associated GRAF Pan-Genome References, first announced in June 2020 and named #3 on The Scientist magazine’s Top 10 Innovations of 2020.
Bioinformatics methods suffer from a significant loss of accuracy when used on sequencing data from under-represented populations because the genetic information of such populations cannot be captured in the linear reference genome. The inherent bias in the industry’s current reference genome favors the European population. A pan-genome approach can provide significant improvements, but it is still suboptimal. The GRAF Population Solution is ancestry-aware and, therefore, more accurate than both linear and pan-genome approaches.
“Precision medicine requires precision bioinformatics methods. Seven Bridges GRAF™ facilitates accurate, comprehensive discovery of genetic variation in the human genome by transforming the existing linear genome reference into a genome graph that incorporates information about the relevant genomic context,” said Brandi Davis-Dusenbery, Ph.D., Chief Scientific Officer, Seven Bridges. “GRAF can construct and utilize genome graphs representing a particular disease, a single individual, a specific population or the entire human species and, therefore, delivers tailored bioinformatics solutions for a multitude of genomics studies.”
The Seven Bridges population-specific NGS analysis pipeline accommodates raw reads in FASTQ, BAM or CRAM format and produces read alignments and variant calls in BAM/CRAM and VCF formats, respectively. When provided with candidate population haplotypes (structural variants), it can accurately genotype the population at scale with respect to those variants. The GRAF Population Solution also improves the accuracy of reads, with a 50% reduction in alignment errors and a 10% increase in variant calling sensitivity. The sensitivity improvements provided by GRAF cover the clinically significant regions of the human genome (exons), in addition to introns and intergenic regions.
In addition to the analysis pipeline, Seven Bridges has developed a robust and effective genome graph construction method and is now offering streamlined development of population-specific genome graphs. The iterative graph construction approach developed by Seven Bridges can augment existing graphs with new data, providing a perpetual improvement cycle for references without compromising compatibility. The five reference population graphs are available for license from Seven Bridges and can serve as baseline references for population-scale research.
“The overarching mission of our GRAF™ Population Solution is to expand the utility of genomics and bioinformatics by enabling science to benefit people of all backgrounds, not just a few that are captured in the current reference standards. GRAF achieves this by constructing ancestry-aware genome graph references that encompass the genetic architecture of the population of interest,” said Dr. H. Serhat Tetikol, GRAF Product Director at Seven Bridges. “Our recent work explores the benefits of a tailored genome graph reference for individuals of African ancestry, which constitute the most diverse and, arguably, the most underserved population.” For information on GRAF, please visit sevenbridges.com/graf.