/mensadb-open

Primary LanguagePython

MENSAdb

The features extracted for the database of membrane protein dimer analysis can be replicated through this repository.

  • INSTALLATION REQUIREMENTS: First you will need to run python setup.py in your terminal to install all the dependencies necessary for feature extraction. Excluded from these dependencies are PSI-Blast and AutoDockTools that you need to install independently. Additionally the non-redundant (nr) database must be downloaded from NCBI (https://ftp.ncbi.nlm.nih.gov/blast/db/).

  • Before feature extraction, you should perform a PRE-PROCESSING of the PDB files. For that you need to:

    • Trim non-transmembrane residues
    • Remove heteroatoms
    • Mutate exotic amino acids
    • Model incomplete structures
    • Dimer extraction from the structure Files
    • Add hydrogens

To see additional details in how to perform data pre-processing, please see our review - "Structural Characterization of Membrane Protein Dimers" published in Methods in Molecular Biology - Protein Supersecondary Structures (https://www.springer.com/us/book/9781493991600).

A - Obtain all the features using a single PDB file as input.

  • run.py deploys all the below features as well as the needed libraries to attain the output files. It will look for information in the intermediate file mensadb_fetcher.py. To attain all the features run: python run.py [pdbid] [chains]

Example: python run.py 1a0t PQ

B - Obtain each feature individually using a single PDB file as input.

  • dssp_features.py extracts the features from a dssp output file. Also requires the corresponding pdb file. To attain the dssp output file use the DSSP executable and run: dssp -i [pdb_name.pdb] >[output_name.txt], in windows, or mkdssp -i [pdb_name.pdb] > [output_name.txt], in UNIX based operating systems. To attain DSSP features, you can run python dssp_features.py, obtaining the following:

    • DSSP index
    • Amino acid number
    • Amino acid code
    • Chain
    • Secondary Structure
    • BP
    • ASA
    • NH-->O_1_relidx
    • O-->NH_1_relidx
    • NH-->O_1_energy
    • O-->NH_1_energy
    • TCO
    • KAPPA
    • Alpha
    • Phi
    • Psi
    • X-CA
    • Y-CA
    • Z-CA

  • features_pssm.py extracts the pssm "jsd" features from psi-blast output file. To retrieve the pssm files needed you will require the psiblast local installation, the non-redundant (nr) database and your input file, with this, run: psiblast -query [fasta_file.fasta] -evalue 0.001 -num_iterations 3 -db [nr] -outfmt 5 -out pssm_output_name.txt -out_ascii_pssm [output_name.pssm] -num_threads 6". Running this step can be very time-consuming, depending on the computer and the protein. To attain PSSM "jsd" features output, you can run: python features_pssm.py.

  • process_binana.py extracts the features from the BINding ANAlyser output file (BINANA - to download go to http://rocce-vm0.ucsd.edu/data/sw/hosted/binana/#download). To attain the BINANA output, you can run: python binana_1_2_0.py -receptor /path/to/receptor.pdbqt -ligand /path/to/ligand.pdbqt -output_file /path/to/output.pdb, as stated in the website of this software. To use this command, you will need their binana_1_2_0.py script, as well as the ".pdbqt" input files. To attain the selected features from the BINANA output, you can run: python process_binana.py. A single csv will be written for each of the possible features. These features are related to a dimer, specifically.

    • Below 2.5 Angstrom residues
    • Below 4 Angstrom residues
    • Hydrogen Bonds
    • Hydrophobic contacts
    • Pi-Pi bond stack
    • T - stack
    • Cation - Pi interaction
    • Salt-bridges

  • generate_class.py uses vmd to extract the interfacial and surface classification for each residue. Makes use of 5 other scripts that are located on the "mensa_class" folder. To use these scripts is required the installation of python based vmd. This can be done with: conda install -c conda-forge vmd-python. The whole code can be run with generate_outputs(input_pdb).joint_call(autodock, autodock_2). Check the path list and replace with your locations. The possible classes are:

    • non-interface and non-surface: 0
    • non-interface and surface: 2
    • interface and surface: 3

References

  • Preto A.J., Matos-Filipe P., Koukos P.I., Renault P., Sousa S.F., Moreira I.S. (2019) Structural Characterization of Membrane Protein Dimers. In: Kister A. (eds) Protein Supersecondary Structures. Methods in Molecular Biology, vol 1958. Humana Press, New York, NY

Please cite

  • Matos-Filipe P., Preto A.J., Koukos P.I., Mourão J., Bonvin A.M.J.J., Moreira I.S. MENSADB: A Thorough Structural Analysis of Membrane Protein Dimers. Available in arXiv:1902.02321 (https://arxiv.org/pdf/1902.02321.pdf)