Skip to content

Using Deep Learning to classify disease-SNPs based on chromosome location

Notifications You must be signed in to change notification settings

bonaventuredossou/deepsnpsthesis

Repository files navigation

Hello, I am Bonaventure Dossou. This is the official repository of my Master's Thesis at Jacobs University Bremen

DeepSNPs: Deep Learning for Single Nucleotide Polymorphism Disease Classification based on Chromosome Location

Raw Datasets

Ready-to-go

Deep Learning Model: Bidirectional LSTM using Keras Tensorflow

  • Overall Workflow Picture

  • Models's weights could be accessed here

  • Results:

    • Training Loss (Accuracy): 0.1947 (0.9016)
    • Validation Loss (Accuracy): 1.5455 (0.7151)
    • Testing Loss (Accuracy): 0.1963 (0.9036)
    • AUC ROC: 0.8733
  • Classification Report

  • Confusion Matrix

  • AUC ROC Curve

  • Features Importance with correct predictions

  • Features Importance with wrong predictions

About

Using Deep Learning to classify disease-SNPs based on chromosome location

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages