Introduction

When Sanger and Coulson first described a reliable, efficient method for DNA sequencing in 1975 (Sanger and Coulson, 1975), they made possible the full sequencing of both genes and entire genomes. Although the method was resource-intensive, many institutions invested in the necessary equipment, and Sanger sequencing remained the standard for the next 30 years.

Refinement of the process increased read lengths from around 25 to almost 750 base pairs (Schadt et al., 2010, fig. 1). Although this greatly increased efficiency and reliability, the Sanger method still required not only large equipment but also significant human investment, as the process requires the work of several people. This prompted researchers and companies such as Applied Biosystems to seek improved sequencing techniques and instruments. Starting in the late 2000s, new instruments came on the market that, although they actually decreased read length, lessened run time and could be operated more easily with fewer human resources (Schadt et al., 2010).

Despite discoveries that have illuminated new therapeutic targets, clarified the role of specific mutations in clinical response, and yielded new methods for diagnosis and predicting prognosis (Chin et al., 2011), the initial promise of genomic data has largely remained unfulfilled so far. The difficulties are numerous. The functional consequences of individual mutations are not always clear. In fact, it is often logistically challenging to determine which discovered mutations make a critical contribution to disease and which are due merely to genetic instability and confer little functional effect.