# Modular Code for Anomaly Detection in Microbiome Data

This repository contains modular Python scripts extracted from a Google Colab notebook implementing an LSTM Autoencoder-based anomaly detection framework for longitudinal microbiome data.

## Files

- `preprocessing.py`: Data loading, cleaning, feature engineering, and sequence construction.
- `model.py`: LSTM Autoencoder model definition and training.
- `evaluation.py`: Anomaly detection, thresholding, and evaluation metrics.

## Usage

Run each script sequentially or integrate them into a pipeline. Ensure required dependencies are installed.

## Contact

For questions, contact: aqureshi.dphd19seecs@seecs.edu.pk