RASEL: Dairy vs Draft Cattle Classifier

This repository contains a trained ensemble model, RASEL ([R-] reference panel, [A-] ancestry informative markers, [S-] selection signatures and the power of [EL-] ensemble machine learning), which classifies cattle populations into Dairy or Draft categories based on SNP genotype data.

Model Overview

RASEL uses an ensemble approach combining:

• Random Forest
• Gradient Boosting
• XGBoost
• Logistic Regression

The model was trained on SNP genotypes from various cattle breeds with known dairy/draft classification. It can be used to classify other cattle populations using the same SNP set.

Repository Structure

├── model/
│   ├── rasel_ensemble_model.joblib          # Trained RASEL ensemble model
│   ├── feature_names.txt                    # SNP names used by the model
│   ├── breed_classes.txt                    # Classification categories (e.g., DR, ML)
│   ├── example_classify.py                  # Example script to use RASEL
│   └── README.md                            # Usage instructions
├── scripts/
│   ├── train_model.py                       # Script used to train RASEL
│   └── evaluate_model.py                    # Script to evaluate RASEL
└── README.md                                # This file

Installation

To use RASEL, install the required packages:

pip install pandas numpy scikit-learn xgboost joblib

Usage

Quick Start

1. Clone the repository:

git clone https://github.com/your-username/rasel-cattle-classifier.git
cd rasel-cattle-classifier

2. Run the example script on your SNP genotype data:

python model/example_classify.py --input your_data.csv --output predictions.csv

Input Data Format

Input CSV should have:

• Column Individuals with sample or breed IDs
• Remaining columns as SNP genotype data (coded as 0, 1, 2)
• SNP columns must match those in model/feature_names.txt

Example structure:

Individuals,BovineHD0100034532,BovineHD4100000740,BovineHD0200004201,...
Sample1,0,1,2,...
Sample2,1,1,0,...

Output

The script outputs:

• Original sample IDs
• Predicted classification (DR for Draft, ML for Dairy)

Citation

If you use this model in your research, please cite:

Kanaka KK, Ganguly I, Singh S, et al. (2025). RASEL: An ensemble model for selection of core SNPs and its application for identification and classification of cattle breeds. Preprint at Research Square. https://doi.org/10.21203/rs.3.rs-6048799/v1

Contact

For support or inquiries, contact: [Kanaka KK](mail to:kkokay07@gmail.com)