sRNA-protein interaction prediction in bacteria

Tahavorgar, Atiyeh (2023) sRNA-protein interaction prediction in bacteria. Masters thesis, Memorial University of Newfoundland.

[img] [English] PDF - Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

Download (1MB)


In bacteria, many biological processes such as stress response, metabolism, and posttranscriptional gene expression regulation are mediated by interactions of proteins with small RNAs (sRNAs). sRNAs are non-coding RNAs (ncRNAs) between 50 to 500 nucleotides long [1]. There are several experimental or wet-lab approaches to determine sRNA-protein interactions; however, wet-lab methods are expensive, timeconsuming, and labor-intensive. Computational approaches, on the other hand, once developed, can predict sRNA-protein interactions quickly and affordably. Current RNA-protein interaction prediction methods have been generated using data from a variety of RNAs (mRNAs, lnRNAs, ncRNAs, etc) and organisms (mammals, bacteria, plants). We hypothesized that a model generated specifically with experimentally validated interacting bacterial sRNA-protein pairs would have a better performance in predicting bacterial sRNA-protein interactions than current methods. To do that, we collected from the literature roughly 1.5k experimentally determined interacting sRNA-protein pairs and used these data to train various machine-learning approaches. Using cross-validation, we selected the most accurate model. Our model achieves an average accuracy of 0.885 ±0.03 on four commonly used RNA-protein interaction data sets which are comparable to other methods. However, we were unable to confirm our initial hypothesis as ProNA’s performance was not better than that of other methods in predicting bacterial sRNA-protein interactions. i

Item Type: Thesis (Masters)
Item ID: 16131
Additional Information: Includes bibliographical references (pages 88-103)
Keywords: sRNA, RNA-binding proteins. machine learning, RNA-protein interaction, bioinformatics
Department(s): Science, Faculty of > Computer Science
Date: August 2023
Date Type: Submission
Digital Object Identifier (DOI):
Library of Congress Subject Heading: Bacteria; Non-coding RNA; Machine learning; Bioinformatics; RNA-protein interactions;

Actions (login required)

View Item View Item


Downloads per month over the past year

View more statistics