circLGB & circMRT

Overview

Identifying circular RNA and predicting its regulatory interactions by machine learning.

Pre-requisite:

• Ubuntu 16.04
• Anaconda 3-5.2.0
• Python packages:
  numpy 1.16.4
  pandas 0.23.0
  scikit-learn 0.19.1
  scipy 1.1.0
  lightgbm

Installation guide

Operation system

Ubuntu 16.04 download from https://www.ubuntu.com/download/desktop

Python and packages

Download Anaconda 3-5.2.0 tarball on https://www.anaconda.com/distribution/#download-section

lightgbm installation:

pip install lightgbm

Content

data

• circLGB: The .bed files for training and testing our circLGB model
• circMRT: The .bed (.fasta) files for training and testing our circMRT model

features

• circLGB: Sequence-derived features of the given sequences for circRNA identification
• circMRT: Sequence-derived features of the query sequences for circRNA regulatory interactions prediction
• feature_importance: Ranked feature list for circLGB and circMRT

weights

• The well-trained weights for circLGB & circMRT models

circLGB.py & circMRT.py

• The Python codes, they can be ran to reproduce our predictors results

result

• The prediction results of our predictors on the test data

Note:

The features/circLGB/test.csv and features/circMRT/test.csv can replaced or modified to the sequence-derived features of interest for identifying circRNA and predicting its regulatory interactions, respectively.

Testing circLGB with test set

python circLGB.py

Testing circMRT with test set

python circMRT.py