To discover rare and common structural variants in a single human sample, we recommend 10-fold coverage as an effective balance of cost and performance. Higher coverage provides improved sensitivity.
|Structural variants (≥50 bp) discovered||18,468|
|Heterozygous variant sensitivity||87%|
|Homozygous variant sensitivity||99%|
|False discovery rate||5%|
*Based on empirical performance of pbsv in a diploid human. Learn more about pbsv here.
To discover common structural variants in a population, we recommend 5-fold coverage per sample and a joint variant caller, which uses reads from one individual to support variant calls in others. More samples provide power to discover rarer variants.
*Modeled performance of a joint variant caller in a population with no substructure and with Poisson sampling of each sample genome.
To discover a causative variant or gene in a disease cohort, we recommend 10-fold coverage for at least three samples with a shared phenotype. Larger cohort sizes provide higher sensitivity for the causative gene and fewer candidate genes for experimental validation.
|Sensitivity for causative gene||99%|
*Modeled performance in a cohort of unrelated individuals each with a structural variant near the same causative disease gene, assuming dominant inheritance.
Please provide your contact details to see experimental requirements and have them emailed to you.