Contaminations contaminate common databases

The polymerase chain reaction (PCR) is a very powerful method to detect and identify pathogens. The high sensitivity of the method, however, comes with a cost; any of the millions of artificial DNA copies generated by PCR can serve as a template in a following experiment. If not identified as contaminations, these may result in erroneous conclusions on the occurrence of the pathogen, thereby inflating estimates of host range and geographic distribution. In the present paper, we evaluate whether several published records of avian haemosporidian parasites, in either unusual host species or geographical regions, might stem from PCR contaminations rather than novel biological findings. The detailed descriptions of these cases are shedding light upon the steps in the work process that might lead to PCR contaminations. By increasing the awareness of this problem, it will aid in developing procedures that keep these to a minimum. The examples in the present paper are from haemosporidians of birds, however the problem of contaminations and suggested actions should apply generally to all kinds of PCR-based identifications, not just of parasites and pathogens.