This is from their conclusions.
Of concern, despite the importance of the actual PCR methods used to amplify these potentially novel viruses, we were unable to determine the type of Taq or master-mix used in 11 of 38 publications. Of the remaining 27 publications, 13 studies used Taq or master-mixes likely containing UNG (it was present in 8 studies and possibly used in the remaining 5 studies that we examined). Therefore contamination at either the individual sample level or in the PCR reaction itself could have inhibited amplification of legitimate target, especially if the target is found at low levels. Additionally, as shown in Figure 2, regardless of the presence of UNG, PCR can be inhibited if there is even a minute amount of contamination from previous (negative) PCR reactions.
Based on our data, we could speculate that low levels of PCR product contamination may not be sufficient to significantly block amplification of the positive control in these reactions, especially if the positive control is the target cloned into a plasmid and present at a high copy number. In the majority of studies commented on in this manuscript, the positive control used was an XMRV plasmid, and most papers were not clear about the amount of plasmid used for control amplification. Therefore, amplification of the high copy number positive control could occur while samples positive for a low copy number virus could be inhibited by carry-over PCR contamination.
