Multiplex Amplicon Libraries

4.6	Multiplex Amplicon Libraries

Multiplex Amplicon libraries are prepared the same way as standard Amplicon libraries with the exception that short Multiplex Identifier sequences, the MIDs, are added to the design of the Adaptors. Since the AVA software expects these sequences at the very beginning of the reads, the MID sequences must be positioned at the end of the Primer A and/or Primer B segments of the Adaptors, just past the sequencing key, and before the template-specific primers (see Figure 4‑7).

Figure 4‑7: Diagrams of potential Amplicon structures using MIDs

(A) MIDs are inserted between the sequencing key and the sequence-specific primers (Primer 1 and Primer 2) on both ends of the Amplicon. A Multiplexer object describing Amplicons with this structure could have an encoding type of “both” (if unambiguous Sample assignment requires both MIDs be found), or “either” (if unambiguous Sample assignment can be made with the MID from either end independently).

(B) An MID is inserted between the sequencing key and the Primer 1 sequence-specific primer only, with no MID used with Primer 2. A Multiplexer object describing Amplicons with this structure must have an encoding type of “Primer 1 MID”.

(C) An MID is inserted between the sequencing key and the Primer 2 sequence-specific primer only, with no MID used with Primer 1. A Multiplexer object describing Amplicons with this structure must have an encoding type of “Primer 2 MID”.

An important difference between the way MIDs are used in Amplicon libraries, compared to Shotgun (sstDNA) libraries, is that Amplicon reads can carry an MID tag at each end, i.e. as part of both Adaptor A and Adaptor B. By contrast, “MID Adaptors” used to prepare Shotgun (sstDNA) libraries, such as the ones provided in the MID Adaptors Kits, carry an MID sequence only on Adaptor A, where the Sequencing Primer of the emPCR Amplification Kit II binds. (Note that the MID kits are not used to prepare Amplicon libraries, since the Adaptors for Amplicon libraries contain template-specific information and must be designed and obtained separately by the user.) The presence of MIDs at both ends of the reads in Amplicon libraries allows their use in a manner analogous to the use of the defined Primer 1 and Primer 2 in the standard (non-MID) AVA demultiplexing scheme; since Amplicons have fully defined sequences (unlike Shotgun library reads), the software knows from the experimental design exactly where the distal MID tag should be, and can thus look for it.

In addition, the possibility to use MIDs at both ends on Amplicons allows for combinatorial demultiplexing, which greatly increases the number of libraries that can be multiplexed with a given set of MIDs. For example, the 454Standard MID Group, which comprises 14 MIDs (see section 1.3.2.6), allows the multiplexing of up to 196 (14 x 14) separate Samples in a single PTP region (Read Data Set), when MIDs are placed at both ends.

However, as different experiments may require different amounts of multiplexing and read length considerations may make it impossible to exploit a demultiplexing scheme in which MIDs are at both ends of the read, the software allows for a number of flexible “encoding” schemes: the user can choose to tag Amplicon libraries at only one end of the reads like in Shotgun libraries (which is simpler), at both ends (to take advantage of combinatorial demultiplexing or to guarantee the ability to demultiplex forward and reverse reads from amplicons that are too large to read all the way through), or at neither end, (i.e. not use MIDs at all and rely on the template-specific Primer sequences to carry out the demultiplexing); encoding schemes are described in more detail in section 1.3.2.7.1.