What is the advantage of unique dual indexing versus combinatorial dual indexing?

Dual indexing in general, which uses both i7 (Index 1) and i5 (Index 2) indices, enables multiple RNA-Seq libraries to be multiplexed in a single sequencing lane or run. Using dual indexing, the combination of both i5 and i7 indices determines the unique barcode combination for sample identification and increases the accuracy of read identification when the sequencing data is demultiplexed.

There are two types of dual indexing: combinatorial dual indexing, and unique dual indexing (Fig. 1).

Figure 1 | Comparison of combinatorial and unique dual indexing strategies. Regardless of the strategy used, each library (A – F) must have a unique combination of i5 (purple, 5001-5006) and i7 indices (yellow, 7001 – 7006). When using combinatorial dual indexing, the same i5 and i7 indices can be used for different libraries as long as the other index is unique (e.g., libraries B and C both have i5 index 5002, but different i7 indices (7002 or 7003)). For truly unique dual indexing, each i5 and i7 index can be used only once, for a single library. In this way only a maximum of 96 uniquely dual-indexed libraries can be multiplexed in a single sequencing lane or run.

Combinatorial dual indexing greatly enhances the multiplexing capacity of a sequencing lane or run. Using Lexogen’s 96 i5 x 96 i7 6 nt indices in combination, up to 9,216 libraries can be uniquely barcoded with different i5 / i7 combinations for multiplexed sequencing in a single run or lane. In doing so, the same i7 and i5 indices can be used for different individual libraries in the pool, so long as the i5 / i7 combination is different for each library.

Unique Dual Indexing differs from combinatorial dual indexing, whereby each i5 and i7 index is present only once in the pool of multiplexed libraries. Unique dual indexing is required to detect index hopping and prevent read mis-assignment during demultiplexing. Index hopping can happen when either the i7 or i5 index of a particular library fragment is switched during cluster generation. Index hopping occurs more frequently on patterned flow cell instruments (HiSeq 3000 / 4000 and NovaSeq) that use ExAmp chemistry, and when excess indexing primers or adapter dimers are present.

Index hopping can also be controlled by repurifying library pools prior to sequencing to remove excess adapter dimers and free index primers. Lexogen’s Purification Module with Magnetic Beads (Cat. No. 022.96) can be used for this repurification step. Briefly, add 0.9 volumes of Purification beads (PB) to the library pool, mix well and incubate for 5 minutes at room temperature, then follow steps 30 onwards in the QuantSeq User Guide (015UG009) Detailed protocol for further details, or contact support@lexogen.com.