A method for constructing a circular RNA-RNA binding protein relationship prediction model

Abstract

The invention discloses a method for constructing a circular RNA-RNA binding protein relationship prediction model, which solves the problem that the prior art only limits the prediction accuracy, and the main point of the technical solution is to construct the initial circular RNA sequence data set sample as In the form of circular RNA-RNA binding protein sequence pairs; self-supervised learning is used to train the word vector dictionary; according to the trained word vector dictionary, the sample sequence pair is mapped into the corresponding word vector matrix as a representation; the sample sequence pair The vector representation corresponding to the input pseudo-Siamese network, the encoded feature vector is obtained and input into the measurement function, the predicted value of the binding probability is calculated and the difference between it and the label is calculated, and the model parameters are optimized; the obtained model is saved after the model training iteration is completed. The method for constructing a prediction model for the relationship between circular RNA-RNA binding proteins of the present invention can perform data mining on RNA sequences and protein sequences, and can effectively improve the accuracy of binding predictions for circular RNA-RNA binding proteins.

Description

technical field [0001] The invention relates to a biological information data mining technology, in particular to a method for constructing a prediction model for the relationship between a circular RNA-RNA binding protein. Background technique [0002] Circular RNA is a special kind of non-coding RNA molecule. Different from traditional linear RNA, circular RNA molecule has a closed ring structure, which is not affected by RNA exonuclease, and its expression is more stable. [0003] Recent studies have shown that circular RNAs play an important regulatory role in diseases and have become the latest research hotspot in the field of RNA. Among them, the research on circular RNA and RNA binding protein (RBP) is a mainstream direction, mainly by studying the regulatory relationship between circular RNA and RBP to better understand the function of circular RNA. [0004] At present, the emergence of a large number of open source high-throughput sequencing experimental data has e...

AI Technical Summary

Problems solved by technology

However, the limitation of the above three methods is that they only use circular RNA sequence information, and judge whether to bind to the target RNA binding protein by learning the specific pattern of the circular RNA sequence, and the binding of the circular RNA sequence to the protein sequence is Realized by the interaction of two sequences, only using circular RNA sequence information limits the prediction accuracy, and there is still room for improvement

Images