Back

SvAnna: efficient and accurate pathogenicity prediction for coding and regulatory structural variants in long-read genome sequencing

Danis, D.; Jacobsen, J. O. B.; Balachandran, P.; Zhu, Q.; Yilmaz, F.; Reese, J.; Haimel, M.; Mungall, C. J.; Beck, C.; Lee, C.; Smedley, D.; Robinson, P. N.

2021-07-15 bioinformatics
10.1101/2021.07.14.452267 bioRxiv
Show abstract

Structural variants (SVs) are implicated in the etiology of Mendelian diseases but have been systematically underascertained owing to limitations of existing technology. Recent technological advances such as long-read sequencing (LRS) enable more comprehensive detection of SVs, but approaches for clinical prioritization of candidate SVs are needed. Existing computational approaches do not specifically target LRS data, thereby missing a substantial proportion of candidate SVs, and do not provide a unified computational model for assessing all types of SVs. Structural Variant Annotation and Analysis (SvAnna) assesses all classes of SV and their intersection with transcripts and regulatory sequences in the context of topologically associating domains, relating predicted effects on gene function with clinical phenotype data. We show with a collection of 182 published case reports with pathogenic SVs that SvAnna places over 90% of pathogenic SVs in the top ten ranks. The interpretable prioritizations provided by SvAnna will facilitate the widespread adoption of LRS in diagnostic genomics.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.