학술논문
The Effects of Normalization, Transformation, and Rarefaction on Clustering of OTU Abundance
Document Type
Conference
Source
2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) Bioinformatics and Biomedicine (BIBM), 2018 IEEE International Conference on. :2810-2812 Dec, 2018
Subject
Language
Abstract
Three common methods, normalization, log transformation, and rarefaction, for handling outliers and erroneous abundances in Operational Taxonomic Unit (OTU) tables are tested for effects in K-means and PCA clustering as well as effects in NMF. We find that of the three methods, log transformation is closest to the unmodified OTUs data, also indicating that when doing specific types of clustering analysis on OTU data, if outliers are to be adjusted within a clustering analysis, a log transformation may be applied.