PriSeT: Efficient De Novo Primer Discovery

Marie Hoffmann; Michael T. Monaghan; Knut Reinert

We evaluated two scenarios for PriSeT. We applied PriSeT on an explicitly uncurated data set (data set for plankton) of short sequences (mostly 18S) taken from large clades to compute primer pairs exhibiting a large taxonomic coverage, and on complete viral genomes (SARS-CoV-2 genomes) to compute primer pairs producing species-distinctive amplicons. As a proof of concept test we additionally searched the primer result set on plankton data for primer pairs used in previous metabarcoding studies. At last, PriSeT's runtime . CC-BY-NC-ND 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.04.06.027961 doi: bioRxiv preprint Algorithm 3 Lookup and encoding of DNA k-mers and references. First the last k-mer occurrence for each reference is determined and a new bit vector instance resized accordingly. Each location L from the input map is associated with a concrete value for k and a list of occurrences represented as tuples of sequence identifiers and positions. The number of bit vector transformed references is referred to as n. In the first loop (line 4-6) bits are set in the bit vector for each kmer occurrence. The second loop (line 8-12) composes the prefix by accumulating the length bits. The third loop (line 13-24) does the sequence lookup for the largest k occurring at a specific position. return Pairs was analized on the plankton data set.