Introduction

RNAcontact is a new algorithm for RNA inter-nucleotide contacts prediction. RNAcontact was built based on deep residual neural networks. The covariance from multiple sequence alignments and the predicted secondary structure were used as the input features. Experiments show that RNAcontact achieves the respective precisions of 0.8 and 0.6 for the top L/10 and L (L is the length of RNA) predictions on an independent test set, significantly higher than other evolutionary coupling methods. Analysis shows that about 1/3 of the correctly predicted contacts are not base pairings of secondary structure, which are critical to decide the shape of RNA structure. In addition, we demonstrate that the predicted contacts could be used as distance restraints to guide RNA structure folding. Significantly more accurate models could be built by using the predicted contacts than without using restraints. Read more about the RNAcontact algorithm...

Submit

Input your RNA sequence in FASTA format. (The length of your input RNA should be in the range of 11~1000 nt.)  Click for an example FASTA input

Or upload the sequence file:

Email: (Optional, where the results will be sent to)

ID: (Optional, your given name to this RNA)

Submit Reset

Reference

S. Sun, W. Wang, Z. Peng, J. Yang, RNA inter-nucleotide 3D closeness prediction by deep residual neural networks, Bioinformatics, 37:1093-1098, 2020.