Study On The Chemical Data Mining Technologies And Their Applications In Drug Design

Along with the development of the modem science and technology,bioinformatics and chemoinformatics become more and more important in the develop of the new drug. The two technologies have become the key factor accelerating the develop of new drug and reducing the cost of developing new drug. It is essential to develop a data mining tool which can help the developer design the new drugs. Visual C++6.0 and OpenGL are used for developing the chemical data mining system in the paper. The key algorithm is Cluster analysis and BP neural networks. This chemical data mining system realizes the visualization of the chemical data mining results,can get the real-time analog display and display of the spatial distribution of the sample point cluster analysis hierarchical graph and genechip. Use the system to do some cluster analysis and BP neural networks patterns recognition experiment with correlative chemical and biologic data used in the develop process of the new drug,and the results show that this mining tool can get the correlative data from the Oracle database easily and do effective data mining,discover and obtain the new knowledge. It is more adaptive to the develop process of drug than many statistical and analytical software in the market at present. In addition,this paper has studied parallel,distributed,multithread computing technology to be realized in Client/Server mode using LAN to have done some tentative work on fast computing tasks distributed in the networks and effective data mining software. And the result of the experiment shows that under the same workload,compared with stand-alone single-thread processing,the p arallel d istributed and multithread c omputing has m ore advantages and can enhance the speed and efficiency of the system,and make computer resource in the network to be more efficiently.