Certainly new algorithms and data structures are going to be crucial (e.g. Bowtie), but statistical analysis/scoring, even in a heuristic way, is always going to be an essential component just from the nature of the work being mostly about evaluating hypotheses from evidence.