Machine Learning Model for Domain Based Classification of MMP’s
The Internet Journal of Genomics and Proteomics · 2010
DOI: 10.5580/1af6 ↗At a glance
Journal article in The Internet Journal of Genomics and Proteomics, 2010. Cited 1 times so far (Crossref).
Abstract (original language)
Purpose
Classification of human matrix metalloproteinase’s (MMP’s or mmp’s) using various machine learning techniques so as to reduce the time and economic constraint which protein classification poses on the existing wet lab techniques.
Methods
The domain information of various MMP’s was obtained from UniProtKB/Swiss-Prot of ExPASy Proteomics Server. Various machine learning tools like Naive Byes, Random Forest and Decision tree were used. The domain data for all the eight classes was incorporated into the three classifiers.
Results
Out of the three the Naive Bayes and Random forest performed best and gave accuracy of 78.26% and 73.91% where out of 23 MMP’s taken for cross validation 18 were correctly classified by Naive Bayes classifier.Discussion and Conclusion: The above classifier takes into account the domain structure of all the known 23 MMP’s as the Naive Bayes classifier performs the best among all the three classifiers, it qualifies as most suitable choice for classification. Information about substrate specificity and tissue inhibitor of matrix metalloproteinase’s (TIMP) was not included since information about these is not completely available. No such algorithm for classification of MMP’s has been reported so far in the literature. Such classification can be extended to other proteins as well where adequate domain information is available.
Cite
APA
(2010). Machine Learning Model for Domain Based Classification of MMP’s. The Internet Journal of Genomics and Proteomics. https://doi.org/10.5580/1af6
Vancouver
Machine Learning Model for Domain Based Classification of MMP’s. The Internet Journal of Genomics and Proteomics. 2010. doi:10.5580/1af6
BibTeX
@article{euvimed2010,
title = {Machine Learning Model for Domain Based Classification of MMP’s},
journal = {The Internet Journal of Genomics and Proteomics},
year = {2010},
doi = {10.5580/1af6}
}
EUVIMED