PDL1 Carbonylation Site Prediction

The repo presents the pipeline of preprocessing protein sequences and predicting protein carbonylation sites.

1. Cut fasta into windows

$ python 1_cut_fasta.py 

2. Compute PSSM features

# install mmseqs
$ wget https://mmseqs.com/latest/mmseqs-linux-avx2.tar.gz
$ tar xzf mmseqs-linux-avx2.tar.gz
$ export PATH=$PWD/mmseqs/bin:$PATH
$ mmseqs databases UniProtKB/Swiss-Prot swissprot tmp

# convert a FASTA file to PSSM
$ sh 2_compute_pssm.sh 

3. Cut PSSM features

$ python 3_extract_pssm.py 

4. Compute PC features

$ python 4_compute_PCfeas.py 

5. Move all files to SCANS folder

$ python 5_move_to_SCANS.py 

6. Run SCANS

# install SCANS
$ wget https://github.com/jianzhang-xynu/SCANS/archive/refs/heads/master.zip
$ pip install fair-esm
$ export PYTHONPATH=/home/rub2/.local/lib/python3.10/site-packages:$PYTHONPATH

$ sh 6_run_SCANS.sh 

7. Combine all sequences

$ python 7_combine_results.py 

8. Plot

$ R 8_plot.R 

Appendix

rm SCANS-master/source_code/SCANS_*/feas/*/*@*
rm SCANS-master/source_code/SCANS_*/seqs/*@*

rm -rf preprocess/PCfeas
rm -rf preprocess/profile
rm -rf preprocess/pssm
rm -rf preprocess/PSSMfeas
rm -rf preprocess/query
rm -rf preprocess/result
rm -rf preprocess/tmp
rm -rf preprocess/window

# SCANS uses older version of mmseqs (Feb 11, 2020).
# there is tiny difference.

mmseqs createdb swissprot.fasta swissprot
mmseqs createindex swissprot tmp