Assay efficiency and sensitivity
The in vitro assessment of assays was carried out with 2 HT-qPCR
runs (96.96 IFC configuration, total 9,216 reactions). The average SD
from both runs was 1.8% indicating little variation between the two
separate runs (supplementary Table S6 ). The 42 assays had an
average efficiency of 94.5% (R2 ranging from
0.99-1.00), with all assays showing efficiency of 90-102.3%, with the
exception of three assays which were found to be below 90%:Msquinado COI (87.8%), Pmaximus COI1 (84.3%), andMedulis PAPM (77.9 %) (Table 3, supplementary
Table S6 ). Most assays showed high detection probabilities with 39 out
of 42 assays having a LOD of <10 copies/µL and the remaining
three assays at 100 copies/µL, lowering to 10-34 copies/µL when using
the modelled approach (Table 3, supplementary Table
S6 ). As for the Limit of Quantification (LOQ) analysis, the discrete
results ranged from 10-1000 copies/µL, whereas the modelled approach for
the LOQ ranged from 10-387 copies/µL, with 23 out of 42 assays obtaining
a 4.4-fold increase in LOQ by the curve-fitting approach over the
discrete approach (Table 3, supplementary Table S6 ).
The LOQ could not be determined for 15 assays as it is below 10 copies.
There was no evident pattern between lower efficiency, LOQ/LOD and the
observed primer variables. Of 42 assays, 33 assays had a LOQ
<50 copies/µL, four assays had a LOQ between 50-387 copies/µL,
and the five remaining assays had a LOQ <61 copies/µL. When
assessing the quantification based on the Standard curve, it appeared
that some assays with similar efficiency had different Cq-values, even
though the starting quantity was the same for all the targets. After the
removal of the outlier assays the efficiency of different assays showed
a linear relationship with the intercept. This model (p
< 0.005, R2 = 0.7) was used to recalibrate the intercept,
which was then used as an alternative method to accurately quantify the
copy numbers (supplementary Table S7 and supplementary Appendix
S2) .