DeeplyEssential

DeeplyEssential is a Deep neural network for the identification of Essential genes in bacteria species. The dataset used for the learning of the nework contains 30 bacterial species collected from DEG.

Dependency

1. Python 2.7
2. keras==2.1.5
3. numpy==1.14.2
4. pandas==0.22.0
5. scikit-learn==0.19.1
6. tensorflow==1.6.0

GPU

Titan GTX 1080 Ti

Parameters

DeeplyEssential takes 6 parameters

1. Essential gene directory path. The directory contains
   • A essential gene sequence file
   • A essential protein sequence file
   • An gene annotation file
2. Non Essential gene directory path. This directory contains
   • A essential gene sequence file
   • A essential protein sequence file
   • An gene annotation file
3. Clustered gene file path clusted by OrthoMCL (sample given, orthoMCL.txt)
4. Text file containing bacteria species information (sample given, dataset.txt)
5. Experiment option
   • '-gp' for Gram Positive (GP) Dataset
   • '-gn' for Gram Negative (GN) Dataset
   • '-c' for GP + GN Dataset
6. Name of the experiment

Run code

$ python main.py <essential gene dir> <non-essential gene dir> <cluster gene file> <dataset> -c <experiment name>

The dataset are collected from DEG. Update: The dataset from current version of DEG may have been changed. Download the exact data that was used for this project from here. (https://drive.google.com/drive/folders/1zhtTP164Ae6MVHrB7A38z8C48tSe0W83?usp=sharing)

Output

DeeplyEssential generates a report containing experiment name, basic statistics about the dataset and evaluation metics for each iteration of experiment. A sample (sample_output.tab) is provided.