Skip to content

A machine learning workflow for comparing and analysing evolutionary rates of homologous protein domain pairs from CATH in terms of sequence and structure.

Notifications You must be signed in to change notification settings

ElofssonLab/evolutionary_rates

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

evolutionary_rates

Workflow for comparing evolutionary rates of homologous protein domain pairs from CATH in terms of sequence and structure.

run_flow.sh

  1. Cluster the data on 95 % sequence identity using CD-HIT. clustering/cluster_95.sh
  2. Download all pdbids with X-ray crystallography and a resolution of less than 2.6 Å (Xray_res2.6Å.txt) (from pdb website)
  3. Filter out the fasta sequences from the 95 % clustering based on the downloaded pdbids: ./pdbfilter reallybelow95.fa Xray_res2.6Å.txt outputdir Writes the fasta sequences that failed the filter into a file.
  4. Select all H-groups with above two entries remaining after the filter and write to individual folders. ./above2.py --H_groups H_groups.tsv --sequences clustering/reallybelow95.fa --failed_pdb_filter (out from previous) --outdir
  5. get_parallel.sh -Convert fasta files to HMMs using hhblits. All pairs of x randomly selected entries are then aligned with hhalign (/get_data.py) -Download the selected pdb files. (download_pdb.sh) -Run DSSP -Run TMalign and tree-puzzle on pdb files (run_tmalign_treepuzzle_ind.py) -Run lDDT on the TMalign alignmens (run_lddt.py) -Run TMscore and tree-puzzle on structural information extracted according to the sequence alignments. -Run lDDT on the structural information extracted according to the sequence alignments.
  6. Merge all results ./merge_results.py --indir INDIR path to input directory. --threshold THRESHOLD Threshold for percent aligned of the shortest sequence in each pair. --outdir OUTDIR path to output directory. --dssp_path DSSP_PATH path to dssp. --fastadir FASTADIR path to fastadir. --gitdir GITDIR path to gitdir. --logfile LOGFILE logfile for stats.

About

A machine learning workflow for comparing and analysing evolutionary rates of homologous protein domain pairs from CATH in terms of sequence and structure.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published