BRCA1 Project: Autoencoder-Based Clustering of Mutation Reads

This repository implements an autoencoder-based pipeline to analyze and cluster BRCA1 mutation reads for haplotype analysis.
The project combines deep learning (autoencoders), clustering, and population genetics data (e.g., gnomAD, MyHeritage raw DNA) to explore the structure and distribution of BRCA1 variants.

---

📌 Overview

• Goal: Detect patterns and outliers in BRCA1 mutations using dimensionality reduction and clustering.
• Methods:
  • Feedforward autoencoder with latent space representation
  • Reconstruction error analysis
  • t-SNE & k-means clustering for visualization
• Applications: Supports genetic research, haplotype analysis, and variant interpretation.

---

⚙️ Installation

Clone this repository:

git clone https://github.com/annadiack/BRCA1-Project.git
cd BRCA1-Project