Evaluation of subsampling-based normalization strategies for tagged high-throughput sequencing data sets from gut microbiomes

Abstract

Several subsampling-based normalization strategies were applied to different high-throughput sequencing data sets originating from human and murine gut environments. Their effects on the data sets' characteristics and normalization efficiencies, as measured by several β-diversity metrics, were compared. For both data sets, subsampling to the median rather than the minimum number appeared to improve the analysis.

Fig. 1.

Effect of the different normalization strategies (color lines) applied on resolution. y axis, average Rt value (the lower the number, the greater the resolution); x axis, rarefaction depth (1, initial; 2, 75%; 3, 50%; 4, 25%; 5, Min). Panel label represents metric used and data set origin. Plots are drawn as line graphs, and y axes are focused to help emphasize the trends across different subsampling depths and strategies.