Advancing analytical algorithms and pipelines for billions of microbial sequences

Abstract

The vast number of microbial sequences resulting from sequencing efforts using new technologies require us to re-assess currently available analysis methodologies and tools. Here we describe trends in the development and distribution of software for analyzing microbial sequence data. We then focus on one widely used set of methods, dimensionality reduction techniques, which allow users to summarize and compare these vast datasets. We conclude by emphasizing the utility of formal software engineering methods for the development of computational biology tools, and the need for new algorithms for comparing microbial communities. Such large-scale comparisons will allow us to fulfill the dream of rapid integration and comparison of microbial sequence data sets, in a replicable analytical environment, in order to describe the microbial world we inhabit.

Figure 1. Moving away from pie charts and trees, current analytical methods

With a few sequences from a small number of samples, pie charts and trees was sufficient for comparing microbial community samples. In contrast, with modern technologies that allow sequencing large number of samples with millions of reads, the new analysis “gold-standard” has moved towards deploying new tools. Here we show data from ref. (6) analyzed with several methods: TopiaryExplorer (41) allows visualization of large trees in the context of per-sample data, , in this example visualizing the GreenGenes reference tree colored by body site matches (red-stool, blue-oral, orange-skin), showing pie charts of most abundant sequences and zooming into the different clades; QIIME PCoA plot comparing all samples color by body site (same colors than in TopiaryExplorer), PCoA with explicit time axis and tracing to follow individuals over time in each body site, allowing visually inspection the changes over time (female: red-gut, blue-oral, orange-skin; male: green-gut, purple-oral, yellow-skin); and semivariograms to assess temporal correlation of observations in the stool samples separated by sex (red-female, blue-male).