3.2. PHASE 2 – IN SILICO SELECTION AND IN VITROASSESSMENT OF MOLECULAR ASSAYS
In total 23 different primer pairs were sourced from the literature (14 publications) and 62 additional primer pairs were designed and selected for further in silico testing (total n = 85 primer pairs). Following initial assessment of each primer pair, 42 assays were selected for further in vitro testing, including 34 assays targeting mitochondrial DNA 16s (n = 3) and COI (n = 31) genes. To gather more information on gene variation and increased taxonomic resolution for Mytilus spp and Magallana gigas , three cytoplasmic organelle 18s gene assays (n = 3) as well as nuclear genes (n = 5) were added to the selected assays (overview of all assays insupplementary Table S4 ).
Results from the initial panel of 42 assays showed that average amplicon length was 104 bp and average primer length was 22.5 bp. The average theoretical annealing temperature was 56.2°C, ranging from 52.7 to 61.7°C, with 17/83 primers outside of the 55-60°C range. Delta-G values of the whole panel averaged -3.9, however, 49/83 primers had one or more reported Delta-G value higher than -6.0 with 26/83 primers having a higher value than -9.0. When blasting the primers, 62/83 were found to match only with the target species. When taking both primers into account, 5 out of the 42 primer pairs matched some non-European off-target species (supplementary Table S4 ). Two primer pairs matched with European species: ‘’M. gigas COI 2” matched withC. angulata (not occurring in Irish coastal waters) and ‘’O. edulis 16S” matched with different Ostrea species and was therefore considered as an Ostrea specific marker. Using the machine learning approach for predicting cross-amplification (eDNAassay), 32 out of 42 assays were found to have off-target AMP values below 0.5 including the highly specific nuclear Mytilus spp markers, hence indicating high specificity potential for 71% of selected assays (supplementary Table S5) . The three nuclear assays targeting the Mytilus complex were Mytilus -specific but had AMP values ranging from 0.5 to 0.675. The results from the COI showed eight assays to have high AMP (>0.5) with off-target species, including five Ensis -specific assays (possible cross-amplification with other Ensis/Pharidae species; 0.5-0.9 AMP), the C. glaucum assay (possible cross amplification withParvicardium exiquum (0.8 AMP)), the Cancer pagurus assay (possible cross amplification with Liocarcinus depurator (0.6 AMP)) and the Homarus americanus assay (possible cross amplification with Eurynome spinosa (0.6 AMP)). Despite the potential for cross-amplification, all 42 assays were tested in vitro , as described below.