grn_inference

Introduction

Currently the tasks include: GRN(Gene Regulatory Network, based on DeepSEM-master)

Installion

Please use conda to create your environment, since we have different models and methods, we recommand to create different envirionments for each cases, here are .yml files as shown in below:

For model GeneCompass(including generate embedding)

# Please confirm prefix in GeneCompass.yaml first !!

conda env create -f GeneCompass.yaml

Pretrained-weights

The Pretrained-weights are provided in "./model/GeneCompass"

How to use this project

Test performence of large model on GRN task

Please run command as shown in below

python get_result-GRN-GeneCompass.py    # For GeneCompass, activate GeneCompass environment

References

For GRN task, we refer DeepSEM framework to complete our jobs, the paper is Modeling gene regulatory networks using neural network architectures. The code is at:(https://github.com/HantaoShu/DeepSEM)