Download IRCAS

Get the source code, datasets, and documentation for IRCAS.

GitHub Repository

Version: 1.0.0

Last updated: April 15, 2024

License: MIT

View on GitHub
Download Stats

Total Downloads: 1,247

This Month: 342

Source Code Size: 85 MB

Dataset Size: 2.3 GB

Support

Documentation: Included in package

Issue Tracker: GitHub Issues

Email Support: Available

Citation: See documentation

Source Code

Complete implementation of IRCAS including all three modules: identification, rectification, and classification.

Contents:
  • main.py - Main entry point 12 KB
  • identification/ - AS detection module 15 MB
  • rectification/ - CNN rectification model 28 MB
  • classification/ - Hybrid GNN classifier 32 MB
  • utils/ - Utility functions 8 MB
  • requirements.txt - Dependencies 2 KB
  • README.md - Documentation 15 KB
Download Source Code (85 MB) View on GitHub

Datasets

Complete datasets used for training and evaluation across four species.

Available Datasets:
  • human_dataset.zip - Human AS events 850 MB
  • mouse_dataset.zip - Mouse AS events 620 MB
  • arabidopsis_dataset.zip - Arabidopsis AS events 450 MB
  • rice_dataset.zip - Rice AS events 380 MB
  • dataset_documentation.pdf - Dataset details 5 MB
  • supplementary_data.docx - Supplementary data 12 MB
Dataset Statistics:
Species AS Types Samples Size
Human 7 190,107 850 MB
Mouse 7 37,924 620 MB
Arabidopsis 7 15,488 450 MB
Rice 7 7,856 380 MB
Download All Datasets (2.3 GB) Download Individual Datasets

Pre-trained Models

Animal-Trained Model

Trained on human dataset, optimized for animal species.

  • Base accuracy: 94.3% (Human), 92.1% (Mouse)
  • Architecture: CNN + GAT + Transformer
  • File size: 125 MB
Download Animal Model
Plant-Trained Model

Trained on Arabidopsis dataset, optimized for plant species.

  • Base accuracy: 96.2% (Arabidopsis), 93.3% (Rice)
  • Architecture: CNN + GAT + Transformer
  • File size: 110 MB
Download Plant Model

Quick Installation Guide

Step 1: Install Dependencies
pip install -r requirements.txt
Step 2: Run IRCAS
python main.py --input transcripts.fasta \
  --output results.csv \
  --species animal
System Requirements
  • Python 3.8+
  • PyTorch 1.10+
  • CUDA 11.0+ (for GPU acceleration)
  • 16GB RAM minimum
  • 50GB free disk space
Quick Start Example
# Download example data
wget http://zhangqblab.cn/IRCAS/examples/sample.fasta

# Run IRCAS
python main.py --input sample.fasta --output results.csv