Skip to content

Latest commit

 

History

History
106 lines (51 loc) · 2.93 KB

README.md

File metadata and controls

106 lines (51 loc) · 2.93 KB

Project Overview

This project focuses on analyzing and predicting the lifespan of worms under various experimental conditions. The primary data consists of CSV files, each of which captures the x-y center-of-mass coordinates of individual worms over time. The project involves processing, feature extraction, and machine learning model training to predict lifespan-related outcomes for worms exposed to different drug conditions.

Data Structure

The data is organized within the ./data/Lifespan directory and follows the structure outlined below.

Subfolders

  • companyDrug: Data from worms exposed to a specific drug.
  • control: Data from worms without drug exposure.
  • Drug2: Data from worms exposed to Drug2.
  • controlDrug2: Data from worms without drug exposure.

Note: The folder names listed above are examples and can be different. You can add additional subfolders to accommodate different drug conditions as needed.

Data Recording Details

  • Time Interval: The time difference between consecutive frames is 2 seconds.
  • Recording Duration: Each recording session lasts for 900 frames (30 minutes).
  • Session Gap: There is a 5.5-hour gap between consecutive recording sessions.
  • Frame Numbering: Frame numbering restarts after 10799 to facilitate better organization and compatibility with the equipment.

Setup Instructions

1. Install Conda

Ensure that you have Miniconda or Anaconda installed on your system.

2. Create a Conda Environment

Use the provided environment.yml file to create a Conda environment named animal-lifespan-prediction:

conda  env  create  -f  environment.yml

3. Activate the Environment

After the environment is created, activate it using:

conda  activate  animal-lifespan-prediction

4. Run the Pipeline

To execute the data processing and feature extraction, run the following command in the appropriate environment:

python run_pipeline.py --base-dir ./data --config pipeline_config.json

Note: Ensure that the folder structure is properly aligned with the "Data Structure" outlined above.

5. Train the Model

To train the model using the processed data, run the following command:

python train_model.py --input ./data/Lifespan_features --config train_model_config.json

Important Note

Ensure that the JSON configuration file train_model_config.json contains the correct subfolder names under the subdirs key within the data object. These subfolder names must align with the actual folder names present in the ./data/Lifespan directory.

Example JSON configuration:

"data": {

"base_dir": "data/Lifespan_features",

"subdirs": ["control", "Drug2", "controlDrug2", "companyDrug"]

}