Automated Annotation of Protein Features Using Language Models

Author(s): LucianoSphere

Originally published on Towards AI.

At UniProt, ProtNLM is already assisting scientists by connecting protein sequences with descriptions of protein traits in the English language

Figure by the author. Note PRTEIN SEQVENCE is not a typo but just reflects the real possible names of the amino acids (O and U don’t exist).

In a nutshell:- ProtNLM is a protein natural language model that is used by UniProt’s Automatic Annotation pipeline for protein annotation.- The model is trained to predict a protein’s name from its amino acid sequence and is based on a sequence-to-sequence model.- ProtNLM can accurately predict descriptions of protein function directly from a protein’s amino acid sequence.- Part of the model was trained to correlate between amino acid sequences obtained from the UniProt database… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI