What is the purpose of Multiplexers?

2.6.5

Multiplexers are used as a means to help the user avoid unnecessary duplication of effort when entering Project setup information and to help make the computation of Project results more efficient by allowing a consolidation of processing steps. To illustrate this, this section describes an example case whereby the Project is specified with and without the use of Multiplexers.

As the starting point, assume an experiment involving 16 different Samples where the same gene (single Amplicon) is being measured in each Sample, but each Sample has a specific pairing of MIDs at each end so that they can be multiplexed together for sequencing. The MID sequences being used are Mid1, Mid2, Mid3, and Mid4 and a ‘both’ encoding is being specified so those 4 sequences can be used combinatorially in pairs to indicate all 16 Samples. All the Samples are multiplexed into a pool and sequenced in two regions of the PicoTiterPlate Device (resulting in two Read Data Sets).

2.6.5.1

Non-Multiplexer Example

First, let’s define the Amplicon to be measured in the 16 different Samples (Figure 2‑47). Since the Samples will be pooled together on the same Read Data Set, MIDs are necessary.

Figure 2‑47: An Amplicon to be measured in 16 different Samples

Using MIDs on both sides of the Amplicon, and assuming the final product is short enough to be sequenced in full within a read, only 4 MID sequences (Figure 2‑48) are needed (combinatorially, using the “Both” encoding) to distinguish the Amplicons from all 16 Samples.

Figure 2‑48: The sequences of 4 MIDs being used to identify 16 Samples by employing a “Both” encoding

If Multiplexers were unavailable it would be necessary to define 16 different Amplicons. This is because in actuality there are 16 different Amplicon library products involved where the MID sequences would need to be considered as part of the template-specific portion of the Amplicon primers. Figure 2‑49 shows the 16 different Amplicons that would need to be created for the experiment. The Amplicons are named according to their MID layout; “Amp_4_3” has Mid4 upstream of Primer1 and Mid3 upstream of Primer 2.

Figure 2‑49: A table of all 16 Amplicons where the MID sequences have been incorporated into the template-specific Primer 1 and Primer 2 sequences. In the highlighted “Amp_1_1” sequence, both primers begin with “ACGAGTGCGT” which is the MID1 sequence from Figure 2‑48 . Since the MID sequences are not actually present in the Reference, the Reference is constrained to the Amplicon being measured so that a single Reference could be used for all of the Amplicons. Otherwise 16 different MID-containing References would have to have been defined.

To finish the setup for a non-Multiplexer experiment, each of the 16 different Amplicons would have to be individually associated with its proper Sample, and those 16 Sample-Amplicon pairs would have to be associated with the Read Data Sets (Figure 2‑50).

Figure 2‑50: Read Data Tree without Multiplexers. Each of the 16 Amplicons had to be individually assigned to a separate Sample and the Sample-Amplicons had to be assigned to the Read Data Sets.

2.6.5.2

Multiplexer Example

With the introduction of Multiplexers, there is no need to define 16 different Amplicons. Only the basic Amplicon in Figure 2‑47 needs to be defined and the Multiplexer contains the information necessary to assign the reads to their proper Sample based on their MID content. This experiment only requires a single Multiplexer that can be used on both Read Data sets. The Multiplexer needs to have the “Both” encoding with 4 MID choices (Mid1, Mid2, Mid3, and Mid4) for Primer 1 MIDs and the same four choices for Primer 2 MIDs. The Multiplexer definition table is shown in Figure 2‑51.

Figure 2‑51: Multiplexer definition table entry

This Multiplexer setup provides a 16 cell grid of Primer 1 – Primer 2 MID pairs that can be assigned to the appropriate Samples (Figure 2‑52).

Figure 2‑52: The Edit Samples window for the Multiplexer. The “Both” encoding being used allows all 16 cells in the MID grid to be assigned to distinct Samples.

To finish off the setup using Multiplexers, the Multiplexer has to be associated with the Read Data Sets. The single Amplicon being measured here can be associated with the Read Data Set –Multiplexer pair and the Amplicon automatically gets associated with each of the Samples encoded by the Multiplexer (see Figure 2‑53).

Figure 2‑53: Read Data Tree with Multiplexers. The Multiplexer is associated with each Read Data Set and the single Amplicon gets associated with each Read Data Set – Multiplexer. The Read Data Set – Multiplexer automatically associates the Amplicon with each of the underlying Samples encoded by the Multiplexer.

2.6.5.3

Multiplexer Benefits Summary

Multiplexers help prevent redundant data entry during project setup when using MIDs. Without Multiplexers, every sample would have to have its own Sample-specific Amplicon defined, and all of those Amplicons would contain some level of duplication of MID sequences contained within them. With Multiplexers, the MID sequences only need to entered once, and one can define the common portion of the Amplicon library product as a single Amplicon rather than needing to define as many Amplicons as there are Samples. Since the Multiplexers also contain the rules for associating an Amplicon with its proper Sample, there is no need to manually make individual Sample – Amplicon associations prior to associating the Sample with the Read Data Set. Instead, the Multiplexer gets associated with the Read Data set and associating the Amplicon with the Read Data Set – Multiplexer pair automatically generates the Sample – Amplicon relationships.

Beyond streamlined data entry, Multiplexers are also important for computational efficiency behind the scenes. The non-Multiplexer example provided in section 2.6.5.1 was included as an illustrative point, but it would run into trouble from a computational point of view. The 16 Amplicons only differ by at most 10 bases in each of their primers. When analyzing an individual read without any foreknowledge of MID specifics, the read needs to be compared against 16 very similar Amplicons. Allowing for distributed error in the read matches to the primer regions, it might be difficult to reliably assign a read to its proper Amplicon. With shorter MID sequences, this would be even more of a problem; the common portions of the primers from all of the Amplicons ends up making the differences in the MID regions seem less significant.

Multiplexers allow a read to be compared with expected template-specific primer sequences to first identify the Amplicon of the read. The knowledge of MID content and layout encoded by the Multiplexer allows the MID regions to be considered in a focused manner after the Amplicon assignment has already been established. This is more efficient and more likely to yield unambiguous results.