Searching for supermaximal repeats in large DNA sequences

We study the problem of finding supermaximal repeats in large DNA sequences. For this, we propose an algorithm called SMR which uses an auxiliary index structure (POL), which is derived from and replaces the suffix tree index ST-FD64 [1]. The results of our numerous experiments using the 24 human chromosomes data indicate that SMR outperforms the solution provided as part of the Vmatch [2] software tool. In searching for supermaximal repeats of size at least 10 bases, SMR is twice faster than Vmatch; for a minimum length of 25 bases, SMR is 7 times faster; and for repeats of length at least 200, SMR is about 9 times faster. We also study the cost of POL in terms of time and space requirements.