A method for predicting lncRNA-protein interaction based on a deep learning dual neural network structure

Abstract

The invention relates to a method for predicting lncRNA-protein interaction based on a deep learning double neural network structure. First, LPI feature extraction: first obtain the known lncRNA sequence and protein sequence, and use Pyfeat and BioTriangle to extract the features of lncRNA and protein respectively; then, feature dimensionality reduction: based on principal component analysis (PCA), the original features of lncRNA and protein are respectively Dimensionality reduction, after dimensionality reduction, connect these features into a vector; secondly, establish the LPI prediction framework model: establish a deep learning model with a dual neural network structure composed of FIR network and MLP network; finally, use the dual neural network structure to analyze the unknown lncRNA‑protein pairs are classified. The present invention is less time-consuming and less expensive than using experimental means to detect large-scale lncRNA-protein interactions, and can realize training and testing of multiple data sets, with small prediction deviation, good prediction performance, accurate prediction results, and can Used to find novel lncRNA‑protein association pairs.

Description

A deep learning-based dual neural network structure prediction of lncRNA-protein phase interaction method technical field The invention belongs to the field of systems bioinformatics, and relates to a kind of double neural network structure prediction based on deep learning lncRNA-protein interaction method. Background technique [0002] Over the past few decades, multiple genome analysis studies have shown that noncoding regulatory elements control complex organic body development process. Noncoding elements are often transcribed into noncoding RNAs (ncRNAs), indicating the importance of ncRNAs in organisms. To regulate roles, studies have shown that ncRNAs can regulate many biological activities that are critical to development, differentiation and metabolism have an important impact. Non-coding RNAs longer than 200 nucleotides are called long non-coding RNAs (Long non- coding RNA, lncRNA), and lncRNAs play a role in regulating cell differentiation by binding t...

AI Technical Summary

Problems solved by technology

First, most of these models are trained and tested on a single data set, which may lead to prediction bias when applied to other data sets, which poses a challenge to improve model prediction performance

However, most methods are not applied to new lncRNA-protein association pair prediction

Second, probing large-scale lncRNA-protein interactions experimentally is time-consuming and expensive

Finally, the predictive performance of these algorithms still needs to be improved

Images