Prediction of RNA Methylation Status From Gene Expression Data Using Classification and Regression Methods

Abstract

RNA N6-methyladenosine (m6A) has emerged as an important epigenetic modification for its role in regulating the stability, structure, processing, and translation of RNA. Instability of m6A homeostasis may result in flaws in stem cell regulation, decrease in fertility, and risk of cancer. To this day, experimental detection and quantification of RNA m6A modification are still time-consuming and labor-intensive. There is only a limited number of epitranscriptome samples in existing databases, and a matched RNA methylation profile is not often available for a biological problem of interests. As gene expression data are usually readily available for most biological problems, it could be appealing if we can estimate the RNA methylation status from gene expression data using in silico methods. In this study, we explored the possibility of computational prediction of RNA methylation status from gene expression data using classification and regression methods based on mouse RNA methylation data collected from 73 experimental conditions. Elastic Net-regularized Logistic Regression (ENLR), Support Vector Machine (SVM), and Random Forests (RF) were constructed for classification. Both SVM and RF achieved the best performance with the mean area under the curve (AUC) = 0.84 across samples; SVM had a narrower AUC spread. Gene Site Enrichment Analysis was conducted on those sites selected by ENLR as predictors to access the biological significance of the model. Three functional annotation terms were found statistically significant: phosphoprotein, SRC Homology 3 (SH3) domain, and endoplasmic reticulum. All 3 terms were found to be closely related to m6A pathway. For regression analysis, Elastic Net was implemented, which yielded a mean Pearson correlation coefficient = 0.68 and a mean Spearman correlation coefficient = 0.64. Our exploratory study suggested that gene expression data could be used to construct predictors for m6A methylation status with adequate accuracy. Our work showed for the first time that RNA methylation status may be predicted from the matched gene expression data. This finding may facilitate RNA modification research in various biological contexts when a matched RNA methylation profile is not available, especially in the very early stage of the study.

Recent advances in functional annotation and prediction of the epitranscriptome

“…The first step of functional epitranscriptome prediction is to identify modification sites, either directly from high-throughput sequencing data or by using sequenced-based computational prediction tools. There exist a large number of sequencing technologies and software tools that can serve this purpose, including but not limited to those based on reverse transcription signature [55] , [56] , [57] , [58] , [59] , bisulfite treatment [39] , [60] , [61] , [62] , [63] , [64] , [65] , [66] , antibody [11] , [12] and the primary sequences of RNA molecules [67] , [68] , [69] , [70] , [71] , [72] , [73] , [74] , [75] , [76] , [77] , [78] , [79] , [80] , [81] , [82] , [83] , [84] . In the following paragraph, we cover primarily two most widely used approaches, including site detection from MeRIP-Seq data and sequence-based in silico prediction methods.…”

PSI-MOUSE: Predicting Mouse Pseudouridine Sites From Sequence and Genome-Derived Features

“…With the massive amount of data generated from various types of high-throughput sequencing techniques, many computational methods have been developed to facilitate the research of RNA modification, such as site prediction and data collection works, [49][50][51][52][53] RNA modification-associated genetic variants analysis tools, 54,55 as well as functional annotation tools. [56][57][58][59] In this study, we innovatively represented 38 RNA topological features on the mouse genome, and a prediction framework PSI-MOUSE was built upon them. Compared with existing works that based on sequence-derived features only, PSI-MOUSE achieved a significant improvement in performance accuracy, indicating the successful application of genome-derived features in the mouse Table 4.…”

“…With the massive amount of data generated from various types of high-throughput sequencing techniques, many computational methods have been developed to facilitate the research of RNA modification, such as site prediction and data collection works, [49][50][51][52][53] RNA modification-associated genetic variants analysis tools, 54,55 as well as functional annotation tools. [56][57][58][59] transcriptome. In addition, 3282 experimentally validated mouse Ψ sites with functional annotations were also collected in our work.…”

PIANO: A Web Server for Pseudouridine-Site (Ψ) Identification and Functional Annotation

“…Alternatively, computational efforts may provide a more cost-effective avenue (Chen X, et al, 2017). To date, many computational efforts have been made to facilitate the study of RNA epigenetics (Boccaletto et al, 2017;Chen X, et al, 2017;Xue et al, 2020;Liu et al, 2020) in terms of both experimental data collection and site prediction works. For predictors related to the identification of Y RNA modification, PseUI (He et al, 2018), XG-PseU (Liu et al, 2019), and iRNA-PseU (Chen et al, 2016) allow for prediction of putative Y sites from an RNA sequence, and PPUS (Li Y.H, et al, 2015) can predict the Y sites regulated by a specific pseudouridine synthase.…”