https://github.com/thehemen/genetic-engineering-attribution-challenge

An algorithm that identifies the lab-of-origin for genetically engineered DNA with the highest accuracy level possible.
https://github.com/thehemen/genetic-engineering-attribution-challenge

Last synced: about 6 hours ago
JSON representation

An algorithm that identifies the lab-of-origin for genetically engineered DNA with the highest accuracy level possible.

• Host: GitHub
• URL: https://github.com/thehemen/genetic-engineering-attribution-challenge
• Owner: thehemen
• License: mit
• Created: 2024-02-07T07:38:15.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2024-03-25T12:51:59.000Z (8 months ago)
• Last Synced: 2024-03-25T14:39:49.792Z (8 months ago)
• Language: Jupyter Notebook
• Size: 204 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Genetic Engineering Attribution Challenge

The goal of [Genetic Engineering Attribution Challenge](https://www.drivendata.org/competitions/63/genetic-engineering-attribution/) was to create an algorithm that identifies the lab-of-origin for genetically engineered DNA with the highest accuracy level possible.

## Main features

The benchmark is extended by more N‑Grams of the ACGT‑sequence as input of the FC network.

The main features of this work:

 - extended N-Grams (six-grams),

 - 8 fold cross validation,

 - separate preprocessing and training stages,

 - fully-connected neural network.

## Results

Public score: 0.9315.

Private score: 0.9181.