Continuous Distributed Representation of Biological Sequences for Deep Proteomics and Genomics.
Asgari E, Mofrad MR. · 2015
Read Continuous Distributed Representation of Biological Sequences for Deep Proteomics and Genomics. by Asgari E, Mofrad MR. (2015) 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/europmc-pmc4640716.txt.
Description
Ehsaneddin Asgari and Mohammad R.K. Mofrad presented a 2015 method that turns biological sequences into number vectors for deep learning. Their BioVec system represents a whole protein or gene as one dense vector, learned by a neural network from sequence patterns. The focus was protein vectors, called ProtVec. The paper appeared in PLOS ONE.
They tested the vectors on 324,018 Swiss-Prot proteins from 7,027 families. Family classification reached about ninety-three percent accuracy, beating existing methods. Vectors also separated disordered proteins from ordered ones with near-perfect accuracy in two test sets. Because only the raw sequence is needed, the approach can feed many downstream tasks, from spotting protein interactions to predicting structure, without hand-built features.
Catalog record
| CREATOR | Asgari E, Mofrad MR. |
|---|---|
| YEAR | 2015 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 36 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |