Skip to content

Latest commit

 

History

History
96 lines (78 loc) · 5.63 KB

README.md

File metadata and controls

96 lines (78 loc) · 5.63 KB

Developer notes

This in an overview of the organization of the following folders and notes for future development efforts.

Overall Design Approach

The software is designed with two programming paradigms in mind:

  • Object-Oriented: By using classes, we try to encapsulate the data and its expected behavior, providing a structured way to represent Structures, which can be converted to AWSEM or DCA models and then be used in subsequent analysis.

  • Functional: Functions are used for specific tasks, offering flexibility and modularity. These tasks include filtering alignments, computing energies, and generating visualizations, among others.

We hope that this combined approach will allow a clean way to manage the data flow using the objects, while offering the flexibility of diverting the flow using the functions.

Folder Structure and Functionalities

The project is structured into several directories, each dedicated to specific types of operations:

classes

  • Purpose: Defines core classes for the project.
  • Key Classes:
    • DCA: Represents a protein with DCA information, including methods to calculate energies, frustrations, decoy statistics, and visualizations.
    • AWSEM: Manages AWSEM energy calculations with methods for computing energies, frustrations, decoy statistics, and visualizations
    • Structure: Manages protein structures, handling PDB files, chains, sequences, and distance matrices.
    • Map: Maps between different sequence representations (for example to translate models of different sizes using alignments)
    • Frustratometer: Base class for Potts-model like models like AWSEM and DCA that provides frustration functions.

frustration

  • Purpose: Contains functions related to the calculations of frustration in protein structures.
  • Energy Funtions:
    • compute_mask: Creates a mask to define which residue pairs to include in frustration analysis based on distance and sequence separation.
    • compute_native_energy: Calculates the native energy of a protein sequence using its Potts model and mask, with options to ignore gap residues.
    • compute_sequences_energy: Efficiently computes energies for multiple protein sequences.
  • Decoy Energy Fluctuation Functions:
    • compute_singleresidue_decoy_energy_fluctuation: Calculates energy changes when single residues are mutated.
    • compute_mutational_decoy_energy_fluctuation: Calculates energy changes for pairwise mutations within contact distance.
    • compute_configurational_decoy_energy_fluctuation: Similar to mutational, but approximates effects on non-contacting residues.
    • compute_contact_decoy_energy_fluctuation: Focuses on changes in pairwise interaction energies upon mutations.
  • Frustration Calculation Functions:
    • compute_single_frustration: Quantifies single-residue frustration based on decoy energy fluctuations.
    • compute_pair_frustration: Quantifies pairwise frustration based on decoy energy fluctuations.
  • Other Functions:
    • compute_scores: Computes contact scores using the Frobenius norm of couplings.
    • compute_roc: Generates a receiver operating characteristic (ROC) curve for assessing contact prediction.
    • compute_auc: Calculates the area under the ROC curve.
  • Visualization functions:
    • plot_roc: Visualizes the ROC curve.
    • plot_singleresidue_decoy_energy: Creates a heatmap or clustermap of single-residue decoy energies.
    • write_tcl_script: Generates a Tcl script for visualizing frustration in VMD.
    • call_vmd: Opens a PDB file in VMD and runs a Tcl script for visualization.

dca

  • Purpose: Handles calculations related to Direct Coupling Analysis using both PyDCA and MATLAB.
  • Key Files:
    • pydca.py: Functions for PLMDCA and MFDCa calculations using PyDCA.
    • matlab.py: Functions for computing and loading Potts models using MATLAB scripts.

filter

  • Purpose: Provides functions to filter and process multiple sequence alignments.
  • Key Files:
    • filter_alignment.py: Filters a Stockholm alignment by removing unaligned sections and gaps.
    • filter_alignment_lowmem.py: A memory-efficient version of alignment filtering.

optimization

  • Purpose: Implements optimization algorithms for protein sequences.

align

  • Purpose: Provides functions for generating alignments.
  • Key Files:
    • jackhmmer.py: Runs Jackhmmer to generate alignments from sequences and databases.
    • extract_sequences_from_alignment.py: Extracts full sequences from a Stockholm alignment file using a specified database.

pfam

  • Purpose: Manages operations related to PFAM alignments.
  • Key Files:
    • download_database.py: Downloads and organizes the entire PFAM database locally.
    • get_alignment.py: Retrieves specific alignments from the local PFAM database.
    • download_alignment.py: Downloads a specific PFAM alignment directly from Interpro.

map

  • Purpose: Offers functions for mapping PDB information to PFAM IDs.
  • Key Files:
    • get_pfamID.py: Retrieves the PFAM ID for a given PDB file and chain.

pdb

  • Purpose: Handles operations related to Protein Data Bank (PDB) files.
  • Key Files:
    • download.py: Downloads a PDB file given its ID.
    • get_sequence.py: Extracts the protein sequence from a PDB file.
    • repair_pdb.py: Repairs a PDB file by inserting missing residues and removing heteroatoms.
    • get_distance_matrix.py: Calculates the distance matrix for specified atoms in a PDB file using various methods.

utils

  • Purpose: Utility functions that don't fit anywhere yet
  • Key Files:
    • create_directory.py: Ensures the creation of a directory if it does not exist.