Technological advances in genomic analyses and computing sciences has led to a burst in genomics data. With these advances, there has also been parallel growth in dedicated accelerators for specific genomic analyses. However, biologists are need of a re-configurable machine that can allow them to perform multiple analyses without needing to go for dedicated compute platforms for each analysis. This work addresses the first steps in the design of such a re-configurable machine. We hypothesize that this machine design can consist of some accelerators of computations common across various genomic analyses. This work studies a subset of genomic analyses and identifies such core computations. We further investigate the possibility of further accelerating through a deeper analysis of the computation primitives.