Main content area

Don't make a mista(g)ke: is tag switching an overlooked source of error in amplicon pyrosequencing studies?

Carlsen, Tor, Aas, Anders Bjørnsgaard, Lindner, Daniel, Vrålstad, Trude, Schumacher, Trond, Kauserud, Håvard
Fungal ecology 2012 v.5 no.6 pp. 747-749
DNA, ecologists, fungal communities, fungi, polymerase chain reaction, sequence analysis
High throughput sequencing has become a powerful tool for fungal ecologists to explore the diversity and composition of fungal communities. However, various biases and errors are associated with the new sequencing techniques that must be handled properly. We here provide evidence for a source of error that has not yet been taken into account. During amplicon pyrosequencing we incorporate tags in both ends of the amplicons, which allows us to check for tag coherence after sequencing. In several studies we have observed that a small proportion of the resulting sequences possess novel tag combinations. Our observations cannot be explained by primer contamination or PCR chimaeras. This indicates that some DNA fragments switch tags during laboratory setup. If not controlled for, this will cause numerous false positives in downstream analyses. In most amplicon pyrosequencing studies of fungal communities, amplicons are typically tagged in one end only. We suggest that amplicons should be tagged in both ends before pyrosequencing to control for tag switching.