synet

• This project aimed to create CNN + LSTM models for 6 ranscription factors

Requirements

* keras==2.8.0
* keras-preprocessing==1.1.2
* tensorflow==2.8.0
* scipy=1.7.3
* numpy=1.21.2

setup environment

• The conda env can be created using synet.yml file

Folder structure

• SOY_models

  • WRKY
  • RAV

• ARB_models

  • NAC
  • MYB
  • bHLH
  • ERF

• Each of the the TF directory has 3 directores.

  • Train : Consist of training scripts and data.
  • Models : The models are saved here by the script. However, you may have to update the paths in the training script.
  • Prediction : Consist of prediction scripts