PyCoM Generator

Pipeline for creating the PyCoM database; see:

• Website: https://pycom.brunel.ac.uk/
• Paper: https://doi.org/10.1093/bioinformatics/btae166
• Python Library repo: https://github.com/scdantu/pycom

tldr: For UniProtKB proteins:

• Filter to manage size (Swiss-prot only; seq length ≤ 500)
• Extract annocations (mjolnir db command)
• Pipeline (mjolnir process command):
  • Sequence Alignment with HH-Suite (HH-Blits => HH-Filter)
  • Predict Protein Residue-Residue Contacts / Coevolution matrices with CCMpred

---

This code was written for the creation of the PyCoM database. It was written to run on the Jade2 HPC at Oxford and the Young HPC at UCL.

The resulting project/database, PyCoM, can be found on https://pycom.brunel.ac.uk.

This git repo is also of interest, as it contains the library for interacting with the database created by this: https://github.com/cemiu/pycom.

My work on this project was funded by the Department of Computer Science, Brunel University London.

DATABASE DOWNLOAD

https://pycom.brunel.ac.uk/

Info

Pipeline tools

Uniclust (for hh-suite)

• http://gwdu111.gwdg.de/~compbiol/uniclust/2022_02/

Sequence alignment

• hhsuite https://github.com/soedinglab/hh-suite

Coevolution matrix

• ccmpred https://github.com/soedinglab/CCMpred