Database Of Structurally Derived Protein Features

Question

[samesense]

I have a list of a thousand residues identified on uniprot proteins. I need to gather features of these residues for a classifier. Some features require protein structures. I have a pipeline for finding appropriate structures using the protein model portal, and calculating structure based features (pka, solvent accessibility, disorder, local charge, etc) using published tools. It took a lot of work to scrape the structures and trouble shoot the published tools. I would have rather used pre computed data, but I could not find such a database.

Are you aware of a protein feature database that spans all uniprot proteins? If not, do you know why such a resource has not been built?

Answer 1

[ddofer]

I wrote a feature generation program/script a while ago in python that can do a lot of this, it's very fast and local, so just run it against the fastas of the proteins you want.

(I'm currently working on a much upgraded version).

Here's the source code.

http://www.protonet.cs.huji.ac.il/neuropid/code/index.php

http://www.protonet.cs.huji.ac.il/neuropid/code/local_SLEEK_FeatureGen+_new.py

Dan

(Used in: http://neuropid.cs.huji.ac.il/ and the articles).