diSBPred

A machine-learning framework for predicting disulfide bonds directly from protein sequence.

Authors: Avdesh Mishra, Md Wasi Ul Kabir, Md Tamjidul Hoque

---

Overview

diSBPred predicts intra-protein disulfide bonds from primary sequence. It ships with a benchmark set and scripts to run natively or via Docker.

• Input: text file at Input/input.txt
• Output: result files written to Output/
• Tested OS: Ubuntu 20.04

If you use diSBPred in published work, please cite the paper listed in Citation.

---

Dataset

The Dataset/ directory contains:

• list_1859_Uniprot_Cutoff25_33_Intra_Cutoff25_Combined_Cutoff25_TotalProts1866.txt — UniProt IDs used to benchmark diSBPred.

Tip: use this list to reproduce reported benchmarks or to sanity-check your setup.

---

Quick Start

# 1) Get the code
git clone https://github.com/wasicse/diSBPred.git
cd diSBPred

# 2) Install dependencies (pyenv, Python 3.7.4, Poetry 1.1.13)
./install_dependencies.sh

# 3) Run (reads Input/input.txt, writes to Output/)
./run_diSBPred.sh

---

Installation

Native (recommended for development)

Prereqs

• pyenv (latest)
• Python 3.7.4 (installed via pyenv)
• Poetry 1.1.13

One-liner:

./install_dependencies.sh

This script installs/initializes pyenv, sets Python 3.7.4 locally, and installs project packages with Poetry.

Docker (fastest to try)

Build locally:

docker build -t wasicse/disbpred https://github.com/wasicse/disbpred.git#master

—or pull the prebuilt image:

docker pull wasicse/disbpred:latest

Run (mounts the current repo’s Input/ and Output/):

./run_diSBPred_Docker.sh "$(pwd)/Input/input.txt" Output

---

Running diSBPred

Inputs

• Default input path: Input/input.txt
• Format: one entry per line (see examples shipped in the repo).
• Ensure the file is readable and non-empty before running.

Command

./run_diSBPred.sh

Outputs

• Results are written to the Output/ directory (created if missing).
• Check the console log for the exact output filenames produced by your run.

---

Project Layout

diSBPred/
├─ Dataset/
│  └─ list_1859_Uniprot_..._TotalProts1866.txt
├─ Input/
│  └─ input.txt
├─ Output/                # created/populated after a run
├─ run_diSBPred.sh
├─ run_diSBPred_Docker.sh
├─ install_dependencies.sh
└─ (source and config files)

---

Reproducibility Notes

• Validated on Ubuntu 20.04. Other Linux distros may work but are not guaranteed.
• For strict reproducibility, prefer the Docker workflow.

---

Troubleshooting

Poetry/pyenv not found

• Re-open your shell to refresh PATH, or source your profile:

  source ~/.bashrc  # or ~/.zshrc

Wrong Python version

• Inside the repo:

  pyenv local 3.7.4
  poetry env use 3.7.4
  poetry install

Permission errors on scripts

chmod +x run_diSBPred.sh run_diSBPred_Docker.sh install_dependencies.sh

Docker “permission denied” on mounts

• On Linux, ensure your user can access the repo path and that $(pwd)/Output exists (or let the script create it).

---

Support

Questions/bugs: Md Tamjidul Hoque — thoque@uno.edu

---

Citation

Mishra, A., Kabir, M. W. U., & Hoque, M. T. (2021). diSBPred: A Machine Learning Based Approach for Disulfide Bond Prediction. Computational Biology and Chemistry, 91, 107436. https://doi.org/10.1016/j.compbiolchem.2021.107436

BibTeX

@article{Mishra2021diSBPred,
  title   = {diSBPred: A Machine Learning Based Approach for Disulfide Bond Prediction},
  author  = {Mishra, Avdesh and Kabir, Md Wasi Ul and Hoque, Md Tamjidul},
  journal = {Computational Biology and Chemistry},
  volume  = {91},
  pages   = {107436},
  year    = {2021},
  doi     = {10.1016/j.compbiolchem.2021.107436}
}