The decline in DNA sequencing costs and the increased efficiency of next generation sequencing platforms have led to a flood of raw sequence data that taxes the analytical capacities of the current bioinformatic and statistical tools. The number of microbiome studies and the sequencing depth of each is likely to continue to grow given their tremendous potential to uncover important roles in ecosystem and human health. Microbiome data is highly complex; it is compositional, multi-dimensional, overdispersed, and sparse. For these reasons, the analytical and statistical pipelines used to analyze these data should be carefully chosen to address both experimental design and the intrinsic features of the expected data. Here, I present three case studies to illustrate the use of deep 16S rRNA gene and metagenomic sequence analysis of environmental microbiomes. The first study assessed the impact of triclosan with and without biochar on soil bacterial communities. The second study used 16S rRNA amplicons to examine the effect of a glyphosate-based herbicide on the intestinal bacteria of mice. The last study used a shotgun-metagenomic approach to shed light on the genetic underpinnings of flurbiprofen degradation in a bacterial consortium enriched from active sludge.