AI-for-siRNA-Effect-Prediction

Project Overview

This project aims to predict the effects of siRNA drugs for the World Second AI4Science Competition. The code and environment setup described here are intended to ensure that results can be reproduced consistently.

System Requirements

• Operating System: Ubuntu 20.04
• Python Version: 3.9.17
• Pytorch Version: 1.12.1
• viennarna Version: 2.6.4
• cudNN Version: cudnn7.6.5_0
• CUDA Version: 10.2
• GPU: NVIDIA TITAN RTX

Environment Setup

The environment dependencies are listed in the environment.yaml file. You can create the Conda environment using the following command:

conda env create -f environment.yaml

Environment Dependencies

The environment.yaml file includes:

• Conda Channels: pytorch, defaults
• Dependencies: Various packages including pytorch 1.12.1 with CUDA 10.2, viennarna==2.6.4,numpy==1.24.4, pandas==1.5.3, scikit-learn==1.3.0, and others.

The complete list of dependencies is detailed in the environment.yaml file, which includes package versions and additional pip-installed packages.

File Structure

project
|-- README.md
|-- sirna_prediction_environment.yaml
|-- data
|   |-- external_data
|   |   |-- readme.md
|   |   |-- train_data_aug3.2.csv
|   |   |-- sample_submission_aug3.2.csv
|-- code
|   |-- main.py
|-- submit
|   |-- submit_20240822.csv

Explanation of External Datasets

• Regarding the external dataset, we did not use additional samples; instead, we calculated some derived features based on the existing samples, such as the binding affinity between siRNA and target mRNA.

Random Seed

• To ensure reproducibility, random seeds are set in the code.