This image was generated by AI.
Ever wonder what a biological curator for ChEMBL does? Read on to find out what we were up to on a Monday, Tuesday, Wednesday, Thursday and Friday during the course of a week in early June.
A typical Wednesday in Sally’s week
Since we had a special event this Wednesday on 10th June (see Emma’s next blog article), I’ll talk about the routine work I completed on 17th June instead. UniProt recently published a new release in which roughly 57 million protein entries were retired. Because ChEMBL maps protein targets to UniProt accessions, my task today is to replace obsolete accessions with active ones wherever possible.This isn’t as straightforward as it may sound, though. I frequently have to go back to the source publication to find more information on a given target. For example, I might need to confirm the exact strain that was studied, compare protein lengths or molecular mass, or check if the authors themselves provided a protein accession I can use as a cross-reference. To work out which replacements are correct, I mostly rely on tools like ID mapping, UniRef clusters and/or BLAST.
Once that’s done, I prepare the SQL statements needed to update ChEMBL. Sometimes I can simply replace the old accession. In other cases, I have to create new targets or remap the affected assays to an existing target. Complications may arise when the target is linked to a drug mechanism in ChEMBL, is part of a multi-protein target, or variant sequences are affected. To prevent mistakes, all proposed updates will be reviewed by my colleague Emma before we make any changes. It will probably take us several days to work through the entire list of deprecated accessions, but it’s crucial in order to keep ChEMBL in sync with UniProt.
Salesia Werner
Comments