Somewhat self-interested plug here: consider working in metabolomics as well. Metabolomics is where sequencing was in ~2008. The physics and chemistry are pretty well worked out (though many improvements are surely coming in the same way that 454 gave way to Illumina, PacBio, Nanopore, etc.). The software and computational workflows are truly awful, like hard to describe bad. The company that figures out metaboloimcs well is going to command a much larger market than genomics - genomics tells you what's possible, metabolomics tells you actually what's happening.