Hi,
I've been trying to use the BioPython Bio.SeqUtils.ProtParam module to perform bulk calculations of the instability index of proteins. The only problem with this approach is that the BioPython implementation doesn't allow the amino acid sequence to contain amino acid codes B, J, O, U, X and Z. However, the ExPASy ProParam tool allows any amino acid code to appear in the sequence submitted.
What I'd like to know is if anyone knows how the ExPASy tool handles the amino acid codes for which there is no stability data. I've tried removing the invalid amino acid codes from the sequence, and calculating the instability index based on the modified sequence. Unsurprisingly this didn't work as the instability calculation is based on dipeptides.
The ExPASy documentation is here, and the stability data comes from this paper.
Simon
•
written
by
SimonCB765 •
I assume that ExPASy ProtParam simply ignores dipeptides that contain non-standard amino acids. For example, if a sequence begins:
BALZTED
then:
BA LZ ZT # invalid
AL TE ED # valid
Clearly Bio.SeqUtils.ProtParam cannot handle this, so you would need to write code which extracts dipeptides from the sequence and plugs the valid ones into the equation for instability index.
That said: I probably would not use ProtParam in the first place. It is an old tool (1990 - one of the first computational tools) and is based on very little protein data. For example, the instability index is based on only 44 protein sequences (12 unstable, 32 stable). I doubt that predictions based on so little information have much validity.
Sorry, I forgot to say that I had attempted that as well. It still seems to give different answers. I do agree with you about the usefulness of the instability predictions though. At this stage it's probably more a sense of curiosity/completeness that's leading me to try and work this out.
Log in to answer this question.
