Prediction Evaluation of Gene Ontology Using Support Vector Machine

Abstract

The present state of sequenced programs requires the assignment of gene product functions in a timely, accurate and trustworthy manner. Many approaches to large-scale label designs have been developed. On the other hand, these approaches can only be used on a limited number of sub-sets. Their conclusions are not formalized. On the other hand, such approaches can only be used on a limited number of subsets, as their conclusions are not standardized. Annotation was supplied using Gene Ontology (GO) or categorization of valid or incorrect prediction using Support Vector Machines (SVM). A large database was used to assess the system's overall effectiveness. Reliability prediction was cross-validated organization by organization, yielding an average accuracy of 74% of all test cycles and 80%. The verification results revealed that the predictive efficacy was not dependent on the micro-organism because it could duplicate the high-quality automatic manual annotation. We used our trained categorization method to annotate Xenopuslaevis sequences, and greater than half of the known expressed genome was functionally annotated. We gave more than double the number of contigs with excellent annotations of high brightness compared to the already accessible annotations, and we also allocated a confidence score to each anticipated Gene Ontology (GO).