dicast: a machine learning method for accurate structural variant detection from short-read sequencing data
Nico Alavi, M-Hossein Moeinzadeh, Jakob Hertzberg, Uirá Souto Melo, Lion Ward Al Raei, Paolo Infantino, Maryam Ghareghani, Marco Savarese, Stefan Mundlos, Martin Vingron · 2026
Read dicast: a machine learning method for accurate structural variant detection from short-read sequencing data by Nico Alavi, M-Hossein Moeinzadeh, Jakob Hertzberg, Uirá Souto Melo, Lion Ward Al Raei, Paolo Infantino, Maryam Ghareghani, Marco Savarese, Stefan Mundlos, Martin Vingron (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13584385.txt.
Description
Nico Alavi and co-authors teach a machine to spot big DNA changes in a 2026 paper. Structural variants, large deletions and rearrangements, cause many diseases. Short-read sequencing, the clinic workhorse, finds them poorly.
Their method, dicast, scores each candidate call with alignment and context traits. It trained on a new ground truth built from nine samples across several technologies, with heavy hand curation. Against rival callers and consensus blends, dicast kept high precision while finding far more true variants. In disease cohorts it caught every known harmful variant. It also found a fifth more candidate harmful deletions than consensus methods. The authors pitch it for routine diagnostics.
Catalog record
| CREATOR | Nico Alavi, M-Hossein Moeinzadeh, Jakob Hertzberg, Uirá Souto Melo, Lion Ward Al Raei, Paolo Infantino, Maryam Ghareghani, Marco Savarese, Stefan Mundlos, Martin Vingron |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 90 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |