Identifying N7‐methylguanosine sites by integrating multiple features

• DOI: 10.1002/bip.23480
• Publication Date: 2021-10-28T15:50:39Z
• Authors: Hongliang Zou; Fan Yang; Zhijian Yin
• Keywords: Complement

Recent studies reported that N7-methylguanosine (m7G) plays a vital role in gene expression regulation. As consequence, determining the distribution of m7G is crucial step towards further understanding its biological functions. Although experimental approaches are capable accurately locating sites, they labor-intensive, costly, and time-consuming. Therefore, it necessary to develop more effective robust computational methods replace, or at least complement current methods. In this study, we developed novel sequence-based tool identify RNA sites. model, 22 kinds dinucleotide physicochemical (PC) properties were employed encode sequence. Three types descriptors, including auto-covariance, cross-covariance, discrete wavelet transform adopted extract features from PC matrix. The absolute shrinkage selection operator (LASSO) algorithm was utilized reduce influence irrelevant redundant features. Finally, these selected fed into support vector machine (SVM) for distinguishing non-m7G proposed method significantly outperforms existing predictors across all evaluation metrics. It indicates approach identifying