Quorum planted ( $l, d$ ) motif search (qPMS) is a challenging computational problem in bioinformatics, mainly for the identification of regulatory elements such as transcription factor binding sites in… Click to show full abstract
Quorum planted ($l, d$ ) motif search (qPMS) is a challenging computational problem in bioinformatics, mainly for the identification of regulatory elements such as transcription factor binding sites in DNA sequences. Large DNA datasets play an important role in identifying high-quality ($l, d$ ) motifs, while most existing qPMS algorithms are too time-consuming to complete the calculation of qPMS in a reasonable time. We propose an approximate qPMS algorithm called APMS to deal with large DNA datasets mainly by accelerating neighboring substring search and filtering redundant substrings. Experimental results on them show that APMS can not only identify the implanted ($l, d$ ) motifs, but also run orders of magnitude faster than the state-of-the-art qPMS algorithms. The source code of APMS and the python wrapper for the code are freely available at https://github.com/qyu071/apms.
