Catalytic Element Mining And Modeling Based On Genomic Sequence

The development of sequencing technologies has generated the accumulation of sequence data.High-quality genome annotation is critical to understanding the function and evolution of organisms.The reactions of the entire metabolic network reflect an organism's chemical repertoire,which drive the synthesis and degradation of the essential molecules of life.It is significant for reconstruction of metabolic networks for purposes of metabolic engineering and drug discovery by accurate prediction of enzyme reactions based.Enzymes as the most important group of proteins catalyze most chemical reactions involved in the metabolism of living organisms.Here,we present Bio2Rxn as model to provide putative enzymatic reaction predictions for uncharacterized protein sequences.Bio2Rxn adopts a consensus strategy by incorporating six types of enzyme prediction tools.It allows for the efficient integration of these computational resources to maximize the accuracy and comprehensiveness of enzymatic reaction predictions,which facilitates the characterization of the functional roles of target proteins in metabolism.To compare the performance of the integrated methods with that of the individual tools,we compiled data from the UniProtKB.Then we compared the prediction results with the manual annotation in UniProtKB/Swiss-Prot.For each number of consensus methods,we calculated:precision,recall and F1-score.Compared with the other tools,when a prediction was produced by 2 methods simultaneously,Bio2Rxn have the highest F1-value(0.75).When a prediction was produced by 3 methods simultaneously.Bio2Rxn has good performance with the highest accuracy(0.91).Django is a mature web application framework written for Python.Based on Django,we present Bio2Rxn,a user-friendly web-based platform for the automatic annotation of the biochemical reactions of protein sequences.Bio2Rxn further links the enzyme function prediction with more than 300,000 enzymatic reactions,which were manually curated over the past 9 years from more than 580,000 publications.The result web page visually shows the transformation of compounds in vivo.On this basis,we developed the website platform V2Rxn(http://design.rxnfinder.org/bio2rxn/v2rxn/).V2Rxn encapsulates gene prediction tools and Bio2Rxn model to build the prediction genome sequence workflows,implementthe prediction of enzyme reaction of genome sequence.