Skip to article frontmatterSkip to article content

Prediction

Overview

You need to decompress the example_data.zip file in your preferred folder, e.g., deepsmirud/.

There are three required files.

We display 4 pairs of small molecules and miRNAs to predict their possible regulation types.

1
2
3
4
5
sm	mirna
5743	MIMAT0000539
84093	MIMAT0009203
148124	MIMAT0000098
5757	MIMAT0000066

Python

We access DeepsmirUD by defining the following parameters.

params = {
    'br_fpn': '../../data/deepsmirud/example_data/br_sm_mirna.txt',
    'smile_fpn': '../../data/deepsmirud/example_data/br_smile.txt',
    'fasta_fp': '../../data/deepsmirud/example_data/',
    'method': 'LSTMCNN',
    'model_fp': '../../data/deepsmirud/model/lstmcnn',
    'sv_fpn': '../../data/deepsmirud/example_data/pred.deepsmirud',
}
Command
Output log
import deepsmirud

deepsmirud.predict.sm_mir_regulation_type(
    br_fpn=params['br_fpn'],
    smile_fpn=params['smile_fpn'],
    fasta_fp=params['fasta_fp'],
    method=params['method'],
    model_fp=params['model_fp'],
    sv_fpn=params['sv_fpn'],
)

CLI

DeepsmirUD can also be used in shell. To know how to use, please type

deepsmirud -h

It shows the usage of different parameters.

-m, --method,
    A deep learning method. It can be any below.
    AlexNet | BiRNN | RNN | Seq2Seq |
    CNN | ConvMixer64 | DSConv | LSTMCNN |
    MobileNet | ResNet18 | ResNet50 | SEResNet
-br, --br_fpn, binary relations between small molecules and mirnas
-sm, --smile_fpn, map between small molecule IDs and their smile strings
-mir, --fasta_fp, miRNA fasta file paths
-mf, --model_fp, a model path
-o, --sv_fpn, outputting deepsmirud predictions

You can run it using the following code.

Command
Output
deepsmirud -m LSTMCNN -br ./data/deepsmirud/example_data/br_sm_mirna.txt -sm ./data/deepsmirud/example_data/br_smile.txt -mir ./data/deepsmirud/example_data/ -mf ./data/deepsmirud/model/lstmcnn -o ./data/deepsmirud/out.deepsmirud
References
  1. Sun, J., Ru, J., Ramos-Mucci, L., Qi, F., Chen, Z., Chen, S., Cribbs, A. P., Deng, L., & Wang, X. (2023). DeepsmirUD: Prediction of Regulatory Effects on microRNA Expression Mediated by Small Molecules Using Deep Learning. International Journal of Molecular Sciences, 24(3). 10.3390/ijms24031878