A linear reference genome is not not an accurate description of genetic variants within a population because it is built from a small number of individuals. This leads to inaccurate read alignment against that reference, which in turn biases the discovery of new genetic variants.
This reference bias can be addressed by using a graph structure to achieve a more complete description of genetic variation within a population. Work with small genomic regions containing high sequence or structural diversity has shown that using a graph reference produces better read alignment and improved variant calling.
Through applying graph technology at the whole genome level, the Seven Bridges Graph Genome Apps enable highly accurate read alignment and somatic variant calling for cancer research applications.
The graph reference genome is built using genomes from a wide range of human populations, including data from the 1000 Genomes Project and the Simons Genome Diversity Project Datasets. The differences between genomes are stored for every position in the genome, resulting in a sequence variation graph. Given that human genomes are 99.9% identical, the increase in storage requirements is minimal with the addition of a new genome to the population reference graph, allowing analysis to be carried out at a previously impossible scale.
- GRAF Germline Variant Detection
- Other graph-based apps will be available soon!
You need a Seven Bridges Platform account to be able to access Graph Genome apps. GRAF Germline Variant Detection Workflow is available in the Public Apps gallery and can be accessed similar to any other app in the gallery.
When you run Graph Genome apps on the Seven Bridges Platform, you will be charged per execution.
Running the GRAF Germline Variant Detection Workflow on the paired-end sequencing data of a whole-genome sample with 30x coverage should cost between $4-6 depending on quality of the sample.
Updated over 2 years ago