Implementation of CReF: a Central Residue Fragment-based method to predict approximate 3-D structures of polypeptides by mining the Protein Data Bank (PDB).

./install

The command downloads and builds dependencies. For details, read the installer/installer.py file.

./cref.sh --help

./cref.sh --pdb 1zdd --pymol

By default the resulting PDB is saved to the predictions folder and named predicted_structure.pdb

./cref.sh --sequence DFYFNAI --output predictions/myprotein/

/cref.sh --fasta 1zdd.fasta

The config.py file contains the default configuration. You can change parameters and specify it on the command line:

./cref.sh --config config --pdb 1zdd --pymol

Note you MUST NOT include the .py in the command line.

The BLAST parameters are used with the ncbi-blast tool. See https://www.ncbi.nlm.nih.gov/books/NBK279675/ for more information.