SPADA Meeting Book

salt concentrations or thermocycling conditions before full-scale assay characterization. Usually 652 a single live organism or inactivated organism strain or extracted/naked DNA/RNA (genomic 653 material) or a surrogate with the assay target or synthetic assay target is used as template at this 654 stage of the assay development process. Minimally, primer concentration optimization, 655 especially if degenerate nucleotides are used, is needed to ensure optimal assay performance. 656 Initial characterization studies can be done in a simple matrix such as water before conducting 657 validation studies in the intended sample matrix. Analytical sensitivity includes determining the 658 assay limit of detection (LOD), or the lowest concentration of organism that is reproducibly 659 detected (e.g., when 58 of 60 test replicates are positive). 660 Analytical specificity involves both inclusivity and cross-reactivity (exclusivity) testing. For 661 inclusivity testing, all the available strains or variants of the targeted organism are tested at 1.5 662 times the LOD. The strains selected for testing should represent any diversity observed within 663 the assay target region. Synthetic nucleic acid could also be used to fill inclusivity gaps once the 664 assay has been optimized and characterized using whole organism. Exclusivity testing should 665 include other organisms located within the target genus, nucleic acid from the intended matrix 666 (e.g., human whole blood), and any organisms that were identified during the initial amplicon 667 BLAST analysis.

Commented [LN(8]: These numbers probably need  additional information to back them up.  I suggest not  providing specific numbers in this example.  It might be  misleading.  Commented [SS9R8]: Koehler question  Commented [LN(10]: We should provide a ref for this  level.  My guess is that other concentrations could be used,  if desired.  Is this from the SMPRs?     Commented [SS11R10]: Koehler question 

668 669

Commented [LN(12]: Was it explained how to do this?  Is  there a cut off to determine which organisms should be  tested?  Or just anything that matches to some level?  Commented [SS13R12]: Koehler; but the general rule of  thumb is 95% match to primer/probe sequences. The  rational is with a couple of mismatches the PCR assay may  work

670

6.2 Assay validation

One approach to validation is described here and Table 2 presents a summary of best 671 practices for PCR assay validation. Assay validation requires a comprehensive analysis of the 672 test system (i.e. the sample to answer process including all steps such as extraction, testing, and 673 analysis) to define the assay performance characteristics in the intended matrix, and the test 674

36