DeepsmirUD

tags: miRNA drugs gene regulation

ℹ️ News!

1. DeepsmirUD is available with being installed with a docker image.
2. Supplementary Tables S5-23 are available in the repository.

Overview

 ____                                _      _   _ ____
|  _ \  ___  ___ _ __  ___ _ __ ___ (_)_ __| | | |  _ \
| | | |/ _ \/ _ \ '_ \/ __| '_ ` _ \| | '__| | | | | | |
| |_| |  __/  __/ |_) \__ \ | | | | | | |  | |_| | |_| |
|____/ \___|\___| .__/|___/_| |_| |_|_|_|   \___/|____/
                |_|

This repository is a standalone package of the DeepsmirUD method. DeepsmirUD is used to predict small molecule-mediated regulatory effects on miRNA expression. This method is powered by 12 cutting-edged deep learning models.

Installation

• PyPI

pip install deepsmirud

• Conda (*python 3.7)

conda install -c jianfeng_sun deepsmirud

• Docker installation

docker pull 2003100127/deepsmirud

Overview

deepsmirud [-h]
           --method m
           --smile_fpn sm
           --fasta_fpn mir
           --model_fp mf
           --output_path o

argument details:
    -h, --help            show this help message and exit
    -m, --method,
            A deep learning method. It can be any below.
            AlexNet | BiRNN | RNN | Seq2Seq |
            CNN | ConvMixer64 | DSConv | LSTMCNN |
            MobileNet | ResNet18 | ResNet50 | SEResNet |


    -sm, --smile_fpn, a small molecule file that contains only smile strings


    -mir, --fasta_fpn, a miRNA fasta file


    -mf, --model_fp, a model path


    -o, --output_path, outputting deepsmirud predictions

Usage

Download models

deepsmirud_download -o /the/path/you/prefer/model.zip

# output messages
downloading...
downloaded!

Please use -mf of deepsmirud then to access to where the models are located.

Input format

Two example files in DeepsmirUD are 5757.txt and MIMAT0000066.fasta for a small molecule and a miRNA molecule.

• Estradiol (small molecule)

# 5757.txt
C[C@]12CC[C@H]3[C@H]([C@@H]1CC[C@@H]2O)CCC4=C3C=CC(=C4)O

• hsa-let-7e-5p (miRNA)

# MIMAT0000066.fasta
>hsa-let-7e-5p MIMAT0000066
UGAGGUAGGAGGUUGUAUAGUU

Inference

Use two example files in DeepsmirUD. The LSTMCNN model is recommanded to use in your studies.

deepsmirud -m LSTMCNN -sm deepsmirud/data/example/5757.txt -mir deepsmirud/data/example/MIMAT0000066.fasta -mf /the/path/you/prefer/model/lstmcnn -o ./out.deepsmirud

Citation

Please cite our work if you use DeepsmirUD in your research.

Contact

If you have any question, please contact Jianfeng Sun. We highly recommend creating issue pages when you have problems. Your issues will be responded then.