Mining Algorithm Design And Implementation Of Web Services-based Biological Information

Web Services is a system framework facing service. The outstanding strongpoint is implemented the platform and language really substantive. This paper issuances the mining algorithm as a service by the Web Services. This paper describes a prototype system that mines biology informations. The two modules were encapsulated in the Web Service. One of the modules was abled to mining the associate rules in gene expression datas and the other one was distinguished the promoters.The mining associate rules module used the FP-tree algorithm in data mining. The experiment testified that FP-tree arithmetic can enhance circulate efficiency. The runtime shortened to 1.045 second, and the originally runtime was 25 hours. The improved promoter prediction module based on the forward algorithm of the Hidden Markov Model. The average distinguish rate of using the purely Markov Model was 84.33%, however the improved system had the rate of 85.17%. The predict-results of the improved Markov Model showed that the system could identify the three sequences effectively. Based on the promoter identification method, this paper worked on the SVM and DNA sequence character based on nonlinear method. Using the long range correlation method in the biological function group RY and structure group MK on second study, was more according with biology characteristic.Web Services technique disposed the biology information by way of mesosphere. They encapsulated arithmetic in Web Services, and only provided the algorithms interfaces. They could integration better functions and had expansibility. The data mining in biology information based on Web Services not only mined association rules which had biological significances in gene expression datas, but also identified promoter sequence. It provided a service platform, which helped the biology and medical affairs communicating with each other simpler and more convenient.