Results and discussion Influence of a single mismatch in the last 4 nucleotides Since the beginning of the 1990s, it has been widely acknowledged that PCR LY294002 molecular weight amplification is significantly inhibited by a single mismatch occurring at the 3′ end of the primer [25–27]. Even when the last nucleotide was substituted with inosine, which is capable of binding to all four nucleotides, primers still failed to amplify all of the expected sequences in the microbial community [28]. Recently, Bru et al. [16] and Wu et al. [17] demonstrated that the efficiency of PCR amplification

was also inhibited if a single mismatch occurred within the last 3–4 nucleotides of the 3′ end of primer, even when the annealing temperature was decreased for optimal efficiency. These single mismatches have not been considered in previous primer coverage studies [12, 18, 29].

We studied the influence of a single primer mismatch occurring within the last 4 nucleotides using the RDP dataset. At the domain level, a relatively weak influence was found when non-coverage rates that allowed a single mismatch in the last 4 nucleotides were compared to rates that did not allow such a mismatch. The absolute differences were ≪5% for all of the primers except 519F (Figure 1A). In contrast, significant differences were observed for some of the primers at the SB202190 concentration phylum level. Rate differences ≫20% under two criteria are listed in Table 1. The most noticeable non-coverage rate was observed for 338F in the phylum Lentisphaerae. If a single mismatch was allowed within the last 4 nucleotides, its non-coverage rate AZD1152 cost was only 3%; otherwise, it was as high as 100%. Similar results were observed for 338F in the phylum OP3, but with a smaller number of sequences. These Chorioepithelioma results indicate that 338F is not appropriate for either phylum (Lentisphaerae or OP3). Overall, the most seriously affected primer was 519F. In this case, 10 phyla showed rate differences ≫20% under two criteria, and 6 phyla showed differences ≫40%. The significant differences observed at the phylum level imply that a single

mismatch in the last 4 nucleotides may be fatal under specific circumstances, and this possibility should be considered when choosing and designing primers. Figure 1 Influence of a single mismatch occurring in the last 4 nucleotides. The black column denotes the non-coverage rate when no mismatches were allowed in the last 4 nucleotides, while the white column denotes the rate when a single mismatch was allowed. A Domain non-coverage rates for 8 primers in the RDP dataset; B Phylum non-coverage rates for primer 338 F in the RDP dataset; C Phylum non-coverage rates for primer 519 F in the RDP dataset. Refer to Additional file 1: Figure S1A for the normalized results of Figure 1A. Table 1 Influence of a single mismatch near the 3′ end in the RDP dataset Primer Phylum Non-coverage rate 4+ (%) Non-coverage rate 4- (%) 338 F Lentisphaerae 3.0 100.0   OP3 5.9 100.