Skip to content
This repository has been archived by the owner on Jun 3, 2022. It is now read-only.
/ deepgene Public archive
forked from yuanyc06/deepgene

Source code of the paper "DeepGene: an advanced cancer type classifier based on deep learning and somatic point mutations"

Notifications You must be signed in to change notification settings

binfnstats/deepgene

 
 

Repository files navigation

DeepGene: An Advanced Cancer Type Classifier Based on Deep Learning and Somatic Point Mutations

This is the software of paper [1]. Please cite [1] if you use this code. Author: Yuchen Yuan Last updated: Oct 29, 2016

Installation

This software is implemented on MatConvNet [2] with CUDA 7.5 and cuDNN v3. CPU-only mode is also supported.

  • Resources: Please download here
  • Supported OS: This software is tested on 64-bit Ubuntu 14.04 and 64-bit Windows 8.1
  • MatConvNet: Please download MatConvNet to the current path, and compile with instructions. Below is a compilation example:
run matlab/vl_setupnn.m
vl_compilenn('enableGpu', true, 'cudaMethod', 'nvcc', ...
'cudaRoot', '/usr/local/cuda-7.5', ...
'enableCudnn', true, 'cudnnRoot', '/usr/local/cuda/');
  • CUDA: If run with GPU, please download and install CUDA
  • cuDNN: If run with GPU, please download and install cuDNN

Usage

  • Entrance: Please run deepgene_demo.m for an example use
  • Default input data path: data/data_mat_all
  • Default trained network path: model
  • Default result file: result.txt
  • GPU or CPU mode: Please set gpus = 1 for GPU mode, or gpus = [] for CPU-only mode.

References

[1] Y. Yuan, Y. Shi et al. "DeepGene: an advanced cancer type classifier based on deep learning and somatic point mutations", BMC Bioinformatics, vol. xx, no. xx, pp. xx-yy, Month. 2016

[2] A. Vedaldi and K. Lenc, "MatConvNet-convolutional neural networks for MATLAB", arXiv preprint arXiv:1412.4564, 2014.

About

Source code of the paper "DeepGene: an advanced cancer type classifier based on deep learning and somatic point mutations"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • MATLAB 26.5%
  • Cuda 21.6%
  • C++ 16.2%
  • Jupyter Notebook 13.4%
  • TeX 9.5%
  • Python 7.5%
  • Other 5.3%