PriSeT: Efficient De Novo Primer Discovery

Marie Hoffmann; Michael T. Monaghan; Knut Reinert

rity thresholds set high (i.e., > 97 %). Despite these natural imperfections, ecologists found that the small subunit RNA gene (SSU) is an excellent source for species-specific markers. Hence a large number of entries in the NCBI-GenBank nucleotide database originate from ribosomal subunits (Benson et al., 2012) . Schmidt et al. (2014) demonstrated that OTUs from 16S/18S rRNA reflect the underlying ecological diversity consistently across habitats. In other words, even if OTUs cannot be resolved, their counts correlate to the sample's diversity and are roughly independent from the species composition. Finding the optimal set of primers in terms of coverage and resolution takes many costly iterations of trial and error (Elbrecht et al., 2019) . Scientists encounter problems like missing ground truth and sparsely populated reference databases. Often they would start with primer sequences published in previous studies on similar data sets and then compare OTU diversity between different primer sets. More formally, primer candidates can be described as kmers -DNA oligomers of length k -satisfying a set of sequence constraints C s which are relevant for the success of the PCR and which must hold independent of their matching. C s comprises range limits for melting temperature, CG content or low probabilities for self-annealing. When combining two k-mers to produce a primer pair, both have to fulfill a second set of constraints C p arising from their strand orientation and fitting: avoidance of CG clamps or low probability of cross-annealing.