mPSQed: a software for the design of multiplex pyrosequencing assays

Abstract

Molecular-based diagnostic assays are the gold standard for infectious diseases today, since they allow a rapid and sensitive identification and typing of various pathogens. While PCR can be designed to be specific for a certain pathogen, a subsequent sequence analysis is frequently required for confirmation or typing. The design of appropriate PCR-based assays is a complex task, especially when conserved discriminating polymorphisms are rare or if the number of types which need to be differentiated is high. One extremely useful but underused method for this purpose is the multiplex pyrosequencing technique. Unfortunately there is no software available to aid researchers in designing multiplex pyrosequencing assays. Here, we present mPSQed (Multiplex PyroSeQuencing EDitor), a program targeted at closing this gap. We also present the design of an exemplarily theoretical assay for the differentiation of human adenovirus types A-F using two pyrosequencing primers on two distinct PCR products, designed quickly and easily using our software.

Figure 1. Principle of multiplex pyrosequencing.

In multiplex pyrosequencing, several primers are used simultaneously in the sequencing reaction so that their signals overlap. A1: In this example, primer 1 (upper part) reads the sequence TTAACCT and primer 1 (middle part) reads the sequence CGCCGTC. Since the signals overlap, the fingerprint (lower part) represents the sequence TTCAAGCCCCGTTC. It is important to note that in this fingerprint, it is not possible to tell which base was read by which primer. A2: The T→C mutation after primer 1 and the C→T mutation after primer 2 are used as targets for differentiating between two species. However, they cancel each other out, causing the fingerprints for A1 and A2 to be identical. B: Moving primer 1 one base to the left alleviates this problem: the fingerprints for B1 and B2 are now different. This demonstrates the importance of correct pyrosequencing primer positioning relative to all utilized SNPs.

Figure 2. View of alignment with consensus sequence displayed for each group.

Zoomed in view of an alignment where groups have been defined. The number of sequences which need to be displayed in order to capture the essential differences between the groups is significantly reduced (the five shown groups contain 94 sequences), but drilling down to the single sequence level is still easily possible, as visible in group “advC”. Bases which are identical to the consensus sequence (or reference sequence, which can be chosen manually) are gray, differing bases are colored based on the selected coloring scheme – “BioEdit” in this case.

Figure 3. Display of SNPs which can be used to differentiate between groups.

SNPs which can be used to differentiate between the defined groups must be perfectly conserved within each group (green column in the group’s consensus graph) and must differ between the groups (orange or red column in the global consensus graph at the top). These positions can be automatically identified and are marked by red columns in the alignment.

Figure 4. Design of multiplex pyrosequencing assay with display of predicted pyrograms.

Display of an alignment with two pyrosequencing primers, one of which is visible on the screen (green annotation), and four PCR primers, one of which is also visible (blue annotation). For each primer, the melting temperature is displayed at the 5′ end and the length is displayed at the 3′ end. A line connects the forward PCR primer with its reverse counterpart (not visible, offscreen). The product size is shown in the middle of the connecting line, and the red color warns of a high difference in predicted melting temperature. In subfigure A, the predicted pyrograms from the two pyrosequencing primers are shown for each group – with just 5 cycles of the pyrosequencing machine, a unique pyrogram can be obtained for each of the groups. In subfigure B, the pyrosequencing primer has been moved one base to the left, thus preventing sequencing of one SNP. This leads to the predicted pyrograms for advE and advB being identical.