HMHO

December 24, 2024 ยท View on GitHub

Heuristic Metropolis-Hastings Optimization Algorithm (HMHO)

Overview

This project provides a Python-based implementation for optimizing protein sequences using Metropolis optimization and external tools like ProteinMPNN and NetSolP. The code allows for the design and analysis of protein sequences, improving their physical properties such as stability and solubility.

Features

  • ProteinMPNN Integration: Generate protein sequences with specific fixed positions.
  • NetSolP Integration: Calculate solubility of protein sequences using a deep learning model.
  • Metropolis Optimization: Iteratively refine sequences to improve biophysical properties.
  • Sequence Analysis: Evaluate sequence instability and flexibility.

Requirements

The project relies on the following dependencies:

Python Packages

Install the required Python packages using:

pip install -r requirements.txt
  • biopython
  • numpy
  • pandas

External Tools

ProteinMPNN

  1. Clone the repository: 
    git clone https://github.com/dauparas/ProteinMPNN.git
  2. Navigate to the directory: 
    cd ProteinMPNN
  3. Install the required dependencies: 
    pip install -r requirements.txt
  4. Verify the setup: 
    python helper_scripts/make_fixed_positions_dict.py --help

NetSolP

  1. Clone the repository: 
    git clone https://github.com/TviNet/NetSolP-1.0.git
  2. Navigate to the directory: 
    cd NetSolP-1.0
  3. Install the required dependencies: 
    pip install -r requirements.txt
  4. Test the setup: 
    python predict.py --help

Usage

Input Files

  1. JSONL Files: Input protein sequence data in JSONL format (e.g., example.jsonl). 
    {"seq_chain_A": "MKRESHKHAEQARRNRLAVALHELASLIPAEWKQQNVSAAPSKATTVEAACRYIRHLQQNGST", "coords_chain_A": {"N_chain_A": [[3.43, -2.059, 57.593], [4.503, -1.273, 55.106], [6.764, -1.617, 53.515], [7.337, -2.021, 50.761], [5.017, -1.711, 49.22], [4.582, 0.937, 49.934], [5.266, 2.436, 47.935], [3.87, 4.624, 46.058], [4.594, 6.744, 47.609], [6.897, 7.564, 46.61], [6.785, 8.744, 43.715], [4.307, 10.33, 44.043], [5.393, 12.123, 45.641], [7.642, 12.655, 44.296], [6.776, 14.022, 42.089], [6.85, 16.764, 42.645], [9.648, 17.311, 42.557], [9.755, 16.909, 39.871], [8.327, 19.099, 38.508], [9.756, 21.05, 39.794], [11.978, 20.909, 38.489], [10.652, 21.715, 36.332], [9.54, 24.523, 36.903], [11.816, 25.563, 36.952], [12.166, 25.521, 34.493], [10.31, 27.34, 33.528], [11.164, 30.041, 34.193], [13.46, 30.237, 32.72], [15.321, 31.184, 29.937], [14.531, 30.564, 26.559], [15.669, 31.934, 24.502], [18.169, 32.237, 25.581], [19.444, 29.701, 25.817], [20.556, 29.18, 23.115], [23.168, 29.865, 23.357], [24.036, 27.974, 25.114], [25.378, 24.724, 24.478], [23.663, 22.355, 22.743], [22.267, 19.861, 24.766], [20.257, 17.991, 25.022], [16.723, 17.849, 25.129], [16.038, 17.345, 28.21], [14.917, 16.05, 31.516], [16.916, 16.592, 33.329], [19.222, 17.178, 31.955], [18.724, 19.942, 31.674], [18.776, 21.005, 34.161], [21.162, 21.433, 34.186], [21.369, 23.694, 32.648], [19.812, 25.819, 34.154], [21.653, 26.638, 35.907], [23.623, 27.828, 34.413], [22.243, 30.231, 33.411], [21.418, 31.434, 35.898], [24.007, 31.44, 37.334], [24.952, 33.312, 35.62], [22.883, 35.722, 35.358], [23.408, 36.895, 37.865], [26.048, 38.607, 37.989], [25.059, 41.093, 36.534], [25.504, 43.588, 34.201], [27.939, 44.523, 35.202], [31.087, 45.887, 34.586]], "CA_chain_A": [[4.785, -2.49, 57.148], [4.43, -0.953, 53.669], [8.138, -1.716, 53.025], [7.329, -2.451, 49.377], [3.839, -1.042, 48.69], [4.622, 2.363, 50.262], [5.438, 2.853, 46.544], [3.362, 5.95, 45.702], [5.331, 7.662, 48.456], [7.945, 8.055, 45.736], [6.296, 9.705, 42.724], [3.357, 11.305, 44.593], [6.304, 13.036, 46.299], [8.597, 12.893, 43.192], [6.091, 14.974, 41.222], [7.375, 18.074, 42.932], [10.944, 17.403, 41.932], [9.527, 16.957, 38.433], [8.003, 20.499, 38.186], [10.825, 21.869, 40.347], [12.772, 20.93, 37.294], [9.795, 22.516, 35.441], [9.476, 25.977, 37.124], [13.199, 25.805, 36.66], [12.029, 25.646, 33.05], [9.492, 28.527, 33.224], [11.937, 31.271, 34.275], [14.724, 30.039, 32.031], [15.133, 31.749, 28.594], [14.848, 29.797, 25.379], [16.602, 32.855, 23.822], [19.478, 32.061, 26.199], [19.787, 28.373, 25.317], [21.384, 29.174, 21.877], [24.514, 30.018, 23.894], [24.356, 26.625, 25.548], [25.587, 23.579, 23.586], [22.494, 21.546, 23.01], [22.533, 18.583, 25.437], [19.044, 17.262, 24.628], [15.665, 18.385, 26.004], [15.771, 16.378, 29.294], [14.569, 16.308, 32.922], [18.19, 16.412, 33.979], [20.084, 17.982, 31.091], [18.411, 21.361, 31.794], [19.126, 21.435, 35.494], [22.474, 21.916, 33.8], [21.081, 25.032, 32.19], [19.394, 26.805, 35.179], [22.872, 27.022, 36.561], [24.139, 28.768, 33.41], [21.42, 31.433, 33.489], [21.611, 31.872, 37.274], [25.419, 31.798, 37.479], [25.009, 34.582, 34.867], [22.017, 36.787, 35.85], [24.082, 37.507, 39.018], [26.84, 39.775, 37.574], [24.412, 41.791, 35.416], [25.928, 44.958, 33.947], [29.31, 44.668, 35.709], [31.966, 46.118, 33.448]], "C_chain_A": [[4.821, -2.46, 55.629], [5.793, -0.89, 52.96], [8.324, -2.273, 51.616], [6.228, -1.644, 48.672], [3.768, 0.46, 48.984], [4.751, 3.145, 48.94], [5.163, 4.307, 46.178], [4.072, 7.086, 46.434], [6.443, 8.292, 47.628], [7.447, 9.139, 44.804], [5.305, 10.744, 43.265], [4.132, 12.414, 45.335], [7.3, 13.547, 45.235], [8.094, 13.988, 42.255], [6.629, 16.385, 41.397], [8.704, 18.185, 42.22], [10.824, 17.491, 40.409], [9.363, 18.427, 37.973], [9.06, 21.456, 38.737], [11.754, 22.018, 39.17], [11.987, 21.813, 36.308], [9.681, 24.04, 35.66], [10.881, 26.508, 36.866], [13.227, 25.974, 35.154], [11.215, 26.888, 32.658], [10.252, 29.876, 33.238], [13.34, 31.114, 33.706], [14.364, 30.54, 30.639], [15.516, 30.908, 27.381], [15.92, 30.62, 24.639], [18.033, 32.781, 24.37], [20.065, 30.837, 25.518], [20.684, 28.245, 24.075], [22.881, 29.246, 22.208], [24.897, 28.616, 24.33], [24.327, 25.548, 24.441], [24.355, 22.72, 23.801], [22.871, 20.178, 23.624], [21.522, 17.575, 24.95], [17.987, 17.825, 25.564], [15.266, 17.39, 27.113], [15.596, 16.887, 30.721], [15.787, 16.105, 33.815], [19.168, 17.329, 33.269], [19.896, 19.52, 31.196], [18.664, 21.906, 33.197], [20.507, 21.998, 35.203], [22.345, 23.394, 33.5], [20.751, 26.095, 33.244], [20.557, 27.376, 35.96], [23.427, 28.135, 35.696], [23.567, 30.174, 33.527], [21.733, 32.135, 34.802], [23.032, 32.337, 37.506], [25.578, 33.143, 36.796], [24.217, 35.781, 35.45], [22.596, 37.563, 37.035], [24.872, 38.769, 38.605], [26.245, 40.508, 36.368], [24.832, 43.267, 35.307], [27.357, 45.322, 34.297], [30.267, 44.848, 34.543], [33.197, 46.98, 33.716]], "O_chain_A": [[5.138, -3.464, 54.967], [5.94, -0.201, 51.946], [9.3, -2.977, 51.346], [6.496, -0.898, 47.738], [3.01, 1.186, 48.316], [4.481, 4.338, 48.88], [6.092, 5.083, 45.958], [4.116, 8.233, 45.962], [6.871, 9.413, 47.898], [7.621, 10.317, 45.099], [5.511, 11.938, 43.065], [3.6, 13.482, 45.616], [7.661, 14.739, 45.221], [8.865, 14.829, 41.788], [6.889, 17.085, 40.43], [8.829, 18.975, 41.283], [11.653, 18.128, 39.735], [10.194, 18.936, 37.191], [9.201, 22.558, 38.243], [12.234, 23.094, 38.865], [12.583, 22.604, 35.572], [9.693, 24.785, 34.673], [11.089, 27.654, 36.439], [14.196, 26.463, 34.609], [11.45, 27.466, 31.594], [10.013, 30.748, 32.392], [14.295, 31.728, 34.181], [13.23, 30.34, 30.188], [16.697, 30.597, 27.172], [17.01, 30.097, 24.342], [18.984, 33.245, 23.724], [21.027, 30.928, 24.766], [21.484, 27.3, 23.999], [23.731, 28.774, 21.441], [25.913, 28.083, 23.887], [23.459, 25.547, 23.532], [24.007, 22.399, 24.943], [23.776, 19.485, 23.142], [21.88, 16.469, 24.57], [18.302, 18.253, 26.684], [14.229, 16.727, 26.998], [16.138, 17.931, 31.102], [15.675, 15.711, 34.964], [19.94, 18.044, 33.892], [20.786, 20.3, 30.81], [18.718, 23.11, 33.416], [20.951, 22.953, 35.845], [22.996, 24.225, 34.135], [21.354, 27.163, 33.226], [20.431, 28.373, 36.669], [23.598, 29.248, 36.164], [24.304, 31.154, 33.674], [22.282, 33.239, 34.803], [23.231, 33.486, 37.9], [26.17, 34.038, 37.38], [24.826, 36.792, 35.865], [22.322, 38.754, 37.166], [24.404, 39.889, 38.848], [26.877, 40.583, 35.302], [24.624, 44.068, 36.237], [27.894, 46.31, 33.764], [30.316, 44.01, 33.633], [33.561, 47.145, 34.904]]}, "name": "1a0aA00", "num_of_chains": 1, "seq": "MKRESHKHAEQARRNRLAVALHELASLIPAEWKQQNVSAAPSKATTVEAACRYIRHLQQNGST"}
  2. CATH Name File: Text file listing JSONL filenames (e.g., cath_name.txt). 
    1a0aA00
1a00B00
1a0cA00
1a0gA02
1a0hA01

Running the Code

  1. Update the paths for ProteinMPNN and NetSolP tools in the script if necessary.
  2. Execute the main script: 
    python protein_optimization.py

Output

  • Optimized Sequences: Stored in updated JSONL files.
  • Logs: Detailed logs of the optimization process.
  • FASTA Files: Generated sequences in FASTA format.

Code Explanation

Main Functions

  1. run_proteinmpnn:
    • Uses ProteinMPNN to generate sequences with fixed positions.
  2. calculate_solubility:
    • Computes solubility using NetSolP.
  3. calculate_instability:
    • Analyzes instability and flexibility of sequences.
  4. metropolis_optimization:
    • Optimizes sequences using a Metropolis algorithm.

Example

Run the optimization pipeline with default settings:

python protein_optimization.py

License

Acknowledgments