🧬 DNA Sequence Classification using Machine Learning

🔹 Overview

This project classifies DNA sequences using machine learning techniques. DNA sequence classification is a fundamental task in bioinformatics with applications in genomics, proteomics, and personalized medicine. This project offers a simple and effective approach to DNA sequence classification using k-mer analysis and the Multinomial Naive Bayes algorithm.

📂 Dataset

The dataset used in this project is sourced from the DNA-Dataset repository. It contains DNA sequences and their corresponding classes, such as human, chimpanzee, dog, etc. The human_data.txt file is used for training and testing the model.

🔬 K-mer Analysis

K-mer analysis involves breaking down DNA sequences into smaller, overlapping substrings of length k (called k-mers).

• Example:
  DNA sequence: ATCGTAC
  6-mers: ATCGTA, TCGTAC

By counting the frequency of different k-mers in a sequence, we can create a feature vector that captures the composition of the sequence.
This approach is effective for DNA sequence classification because different classes of DNA may have distinct k-mer frequency profiles.

⚙️ Dependencies

• pandas
• numpy
• scikit-learn
• matplotlib
• seaborn

You can install these dependencies using pip:

pip install pandas numpy scikit-learn matplotlib seaborn

🚀 Usage

1. Clone this repository (project code):

git clone https://github.com/ZahraSahranavard/DNA-Sequence-ML

cd DNA-Sequence-ML

2. Clone the dataset repository:

git clone https://github.com/ZahraSahranavard/DNA-Dataset.git

3. Run the notebook:

• Open DNA-Data-Analysis.ipynb in Google Colab or Jupyter Notebook.

• Run all cells to train and evaluate the model.

📊 Results

The Multinomial Naive Bayes classifier shows excellent performance on the dataset:

• Accuracy: 0.984
• Precision: 0.984
• Recall: 0.984
• F1-score: 0.984

These results demonstrate that k-mer based feature extraction + Naive Bayes is an effective approach for DNA sequence classification.

🔮 Future Work

• Explore alternative classifiers such as Support Vector Machines (SVM), Random Forest, and Gradient Boosting to compare their performance with Naive Bayes.
• Incorporate DNA-specific embeddings (e.g., DNA2Vec or k-mer embedding models) to capture deeper biological patterns in nucleotide sequences.
• Leverage deep learning architectures like Convolutional Neural Networks (CNNs) for motif detection, Recurrent Neural Networks (RNNs) for sequence modeling, and Transformers for capturing long-range dependencies in DNA.

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.