The original site of the pdf file is http://bioinformatics.oupjournals.org/cgi/screenpdf/18/2/235
A
data-mining approach to spacer oligonucleotide typing of Mycobacterium
tuberculosis
Date
: Mar. 22, 2002
Place
: Conference Room No.1, Activity Hall, NYMU
Speaker :
Pei-Ju Chin
Tuberculosis is the harmful epidemic perplexed human and livestock for thousand of the years. Key factors in the control of tuberculosis are rapid detection, adequate therapy, and contact tracing to arrest further transmission. The Direct Repeat (DR) locus of Mycobacterium bovis is a suitable model to study molecular epidemiology and the evolutionary genetics of tuberculosis. The technique, called spacer oligonucleotide typing (spoligotyping) , existing the characteristics described above. Because of the widely-variant substrain of M. tuberculosis, it becomes a important purpose to handle the huge dataset and when the genotyping data inputed, the system may make the decision properly and rapidly.
The system created in order to accomplish these goals using the following approach: discovering the knowledge rules from spoligotyping, eliminating the noisy data, and finally, verifying the performance and accuracy of the data analyzing model. Surprisingly, comparing with the result which system and human brain generated, the accuracy of system is higher than human. It will become practical and play an important role for control and prediction of the epidemic in the future.
Reference cited