My lab designed a qPCR assay for amplifying the denitrification gene nirS from DNA extracted from soil. The assay has worked perfectly in the past with almost perfect efficiency and gene copy number results. However, recently we noticed that the assay was underestimating by 2 log (Should be at least 10^5, but coming out as 10^3). This was noticed as the gene copy numbers of the external positive control, Pseudomonas aeruginosa, were 2 logs lower than they should be. We tested a number of things to explain this. We regrew and extracted fresh DNA from a new strain of P. aeruginosa, yet the results were the same. BSA was used in this assay, but was also found to not affect gene copy numbers to this extent. Standards and stocks were all re-quantified using Qubit. We are still unsure what is causing the underestimation and how we should to proceed. Does anyone have any suggestions for troubleshooting this issue? Or know of any published research that ran into similar issues? All suggestions are welcome, thank you.