It has been previously reported that ‘homopolymer errors’ (a term originating from Roche 454 pyrosequencing) are the dominant error type in PGM data [2], [3]. Homopolymer errors are a consequence of inaccurate flow-values resulting in over (insertion/s) or under-calling (deletion/s) the length of homopolymeric regions. For consistency with the literature, we refer to these as homopolymer errors, but emphasise that over and under-called flows are not always associated with homopolymers of length two or more [13], illustrated in Figure 4 . This distinction is particularly pertinent for PGM data, where over-calling any flow-position can lead to the insertion of additional bases, even in flows where a ‘negative flow-value’ (<0.50) should have been called.
