Abstract
N6-methyladenosine (m6A) is closely related to various life processes and diseases. The detection of genomic-level m6A sites plays an important role in explaining its biological mechanism. However, the current mainstream m6A sites detection method has limited precision. In this paper, a novel m6A sites detection method called “Dm6A-TSVM” is proposed. The feature vectors are firstly extracted from the mRNA sequences according to their nucleotide chemical property and position statistical distribution characteristics. Then the two kinds of features are combined, and the m6A sites detection model is constructed by the twin support vector machines method. Finally, based on the standard yeast dataset, the cross-validation experimental method is used to verify Dm6A-TSVM. The results demonstrate that the Dm6A-TSVM method is significantly better than the current mainstream m6A sites detection method, and its accuracy (ACC) value reaches 82.81%.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
