README.md

metadata

title: SoluProtMut
emoji: 🔬
colorFrom: red
colorTo: blue
sdk: gradio
sdk_version: 4.44.1
app_file: app.py
pinned: false

Installation

Clone the repository ???.
Then, extract the FoldX distribution for your operating system into external/foldx in the cloned project folder, make sure the binary is named foldx.
Finally, install the dependencies and check the installation works:

pip install -r requirements.in
python -m tests.inference

If the installation fails, or the application crashes, try to install pinned dependencies instead: pip install -r requirements.txt. Also check the table below.

Tested configurations:

OS	Python	PyTorch	Engine	Project version
Win 10	3.8.18	2.1.2	CPU	January 16, 2024
Linux	3.9.19	2.1.2	CPU	January 16, 2024
Linux	3.8.13	1.10.0	CUDA 11.3	January 16, 2024

Usage

wrapper.py [-h] [-v] pdb-code [chain] wild-type location mutation

Solubility change preditor:
  Predicts the change in the solubility of a given protein variant

positional arguments:
  pdb-code       PDB code
  chain          target chain          character; default=A
  wild-type      wild-type residue(s)  amino-acid[n]
  location       mutated position(s)   integer[n]
  mutation       1 or n substituents   amino-acid|amino-acid[n]

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose

Amino acids can be specified by both 1- or 3-letter code.

Example usage (double-point mutant in erythropoietin)`and its outpus:

./wrapper.py 1EER F,R 48,150 D --verbose

Predicted solubility change: 5.000398e-01 (solubilizing)

Development

Reproducible environment

requirements.in – project (direct) dependencies required for inference. Keep the PyTorch version specified here in sync with the link for the pyg wheels specified ibidem.

requirements.txt – resolved and pinned dependencies based on the file above, a result of pip install -r requirements.in && pip freeze in a fresh Python environment.
note: pip-compile of pip-tools does not seem to work with complex PyTorch ecosystem

Style guide

Indentation

Use tabs \t for an indentation, spaces (after tabs) for an alignment (reasoning, reasoning2)

Naming conventions (PEP8-based)

module_name, CONSTANT_NAME, ClassName (capitalize first letters of words), subroutine_name, variable_name

ClassName, function_name: keep acronyms in their original case: GetPDB, get_PDB