This GitHub repository is dedicated to creating the Pfam seeds structure database (PfamSDB) by cutting the related full-length structures at their domain borders.
- aria2c==1.36.0
- Python==3.10.12
- gsutil==5.7 (Access to the AlphaFold database is assumed)
- GNU parallel==20210822
- seqtk==1.4
- Foldseek==8-ef4e960
Python package dependencies are provided inside the requirements.txt file.
- High-speed internet connection
- At least 700GB of storage, preferably on a high-speed SSD
Ensure you have the required software installed. Python libraries can be installed via the requirements.txt file
To create the database of cif files for Pfam seeds, use the make command.
For users who prefer to query protein structures against the PfamSDB using
Foldseek, there is an alternative approach to constructing PfamSDB.
Instead of utilizing cif files of the structures, this method involves
segmenting the Foldseek database along the domain borders.
This alternative technique can be executed using the make pfamsdb_fscut command
To clean up and remove any intermediate files generated during the construction of
PfamSDB, you can use the make clean command.
If you use PfamSDB or any associated tools in your research, please consider citing our work. Here's the citation format:
Borujeni PM, Salavati R. Functional domain annotation by structural similarity. NAR Genomics and Bioinformatics. 2024 Mar 1;6(1).
BibTeX entry for LaTeX users:
@article{borujeni2024functional,
title={Functional domain annotation by structural similarity},
author={Borujeni, Poorya Mirzavand and Salavati, Reza},
journal={NAR Genomics and Bioinformatics},
volume={6},
number={1},
pages={lqae005},
year={2024},
publisher={Oxford University Press}
}