OperonSEQer: A set of machine-learning algorithms with threshold voting for detection of operon pairs using short-read RNA-sequencing data

Abstract

Operon prediction in prokaryotes is critical not only for understanding the regulation of endogenous gene expression, but also for exogenous targeting of genes using newly developed tools such as CRISPR-based gene modulation. A number of methods have used transcriptomics data to predict operons, based on the premise that contiguous genes in an operon will be expressed at similar levels. While promising results have been observed using these methods, most of them do not address uncertainty caused by technical variability between experiments, which is especially relevant when the amount of data available is small. In addition, many existing methods do not provide the flexibility to determine the stringency with which genes should be evaluated for being in an operon pair. We present OperonSEQer, a set of machine learning algorithms that uses the statistic and p-value from a non-parametric analysis of variance test (Kruskal-Wallis) to determine the likelihood that two adjacent genes are expressed from the same RNA molecule. We implement a voting system to allow users to choose the stringency of operon calls depending on whether your priority is high recall or high specificity. In addition, we provide the code so that users can retrain the algorithm and re-establish hyperparameters based on any data they choose, allowing for this method to be expanded as additional data is generated. We show that our approach detects operon pairs that are missed by current methods by comparing our predictions to publicly available long-read sequencing data. OperonSEQer therefore improves on existing methods in terms of accuracy, flexibility, and adaptability.

Influence of genomic variations on glanders serodiagnostic antigens using integrative genomic and transcriptomic approaches

“…Based on the simple model that the destruction of operon structure or disruption of the non-coding region upstream of operons can interfere with gene expression ( 46 ), variant calls (SNPs, INDELs, and SVs) were compared with RNA-seq data to link DEGs to specific genomic variation. Specifically, we identified the operons using operonSEQer v1.0 ( 47 ) and investigated changes in gene expression in relation to an alteration in the operon structure mediated by SVs, as well as changes in the non-coding region upstream of the operon, including promoter and regulators, mediated by SNPs or INDELs. This was achieved using custom scripts modified from a previously described method ( 48 ).…”

Characterization of radiation-resistance mechanism in Spirosoma montaniterrae DY10T in terms of transcriptional regulatory system

“…Operon prediction and clustering. Krishnakumar et al developed a operonSEQer 39 , a state-of-the-art statistic and machine learning based algorithm that predicts bacterial genome operon. operonSEQer shows remarkable predictive power using only RNA-seq data and genomic features.…”

“…It is an important part of understanding the complex regulatory system of bacteria. operonSEQer 39 is a statistics and machine learning based algorithm that predicts relevant operon pair with signals from RNA-seq data across two genes. Unlike other existing tools, operonSEQer is a flexible tool that does not use functional relationship information and showed remarkable operon predictive performance to bacterial strains that have not been studied.…”

Noncontiguous operon atlas for the Staphylococcus aureus genome

“…The set of operons present in a bacterial genome can be computationally predicted with reasonable accuracy based on criteria such as the proximity of neighbouring genes with short intergenic distances across diverse bacterial species and functional relationship between the protein products (Mao et al 2009 , Taboada et al 2012 , 2014 , 2018 , Krishnakumar and Ruffing 2022 ). As these criteria are fully adapted to the classical operon concept, any variation in gene organization that does not meet these criteria is not considered in current operon identification tools.…”