Abstract:N-6-Methyladenosine (m(6)A) is the most prevalent and abundant modification in mRNA that has been linked to many key biological processes. High-throughput experiments have generated m(6)A-peaks across the transcriptome of A. thaliana, but the specific methylated sites were not assigned, which impedes the understanding of m(6)A functions in plants. Therefore, computational prediction of mRNA m(6)A sites becomes emergently important. Here, we present a method to predict the m(6)A sites for A. thaliana mRNA sequence(s). To predict the m(6)A sites of an mRNA sequence, we employed the support vector machine to build a classifier using the features of the positional flanking nucleotide sequence and position-independent k-mer nucleotide spectrum. Our method achieved good performance and was applied to a web server to provide service for the prediction of A. thaliana m(6)A sites. The server also provides a comprehensive database of predicted transcriptome-wide m(6)A sites and curated m(6)A-seq peaks from the literature for query and visualization. The AthMethPre web server is the first web server that provides a user-friendly tool for the prediction and query of A. thaliana mRNA m(6)A sites, which is freely accessible for public use at http://bioinfo.tsinghua.edu.cn/AthMethPre/index.html.

AthMethPre: a web server for the prediction and query of mRNA m(6)A sites in Arabidopsis thaliana