Unique Dual Index for MGI platforms

Background

With continuous upgradation of the next-generation sequencing technology, the HiSeq-series and NovaSeq-series sequencers from illumina, as well as the MGISEQ-series sequencers from MGI Tech, can achieve over hundreds of GB of data at a time. In order to maximize the utilization of the data output, samples are pooled together and performed multi-plex sequencing: namely adding the unique barcode sequence ( or index) into each sample during the library preparation process, and splitting the data in terms of the indexes after sequencing.

The abnormalities occurred in library preparation, sequencing and data analysis may result in the index misassignment, including synthesis error/contamination, improper operation, abnormal amplification, sequencing errorand bioinformatic analysis. Therefore, researchers should highly focused on the possible incorrect conclusions resulted from index misassignment. As a matter of fact, the ExAmp amplification pattern in illumina sequencing platforms have a severe proportion (0.25-7%) of the index misassignment. Fortunately, the introduction of Unique Dual Index (UDI) can greatly reduce the negative impacts caused by the index misassignment^[1].

Due to the PCR-free DNB amplification pattern in MGISEQ-series sequencing platforms, the proportion (0.0004% on average, peak at 0.001%) of single index misassignment is ultralow. Under this condition, 68% of the index misassignment is resulted from the sequencing error,contamination of synthesis and operation^[2]. However, index misassignment may increase when performing targeted capture procedures, which have several steps closely related to the generation of index misassignment^[3].

Compared with UDI strategy in illumina platform, much index hopping contaminations were detected in targeted capture sequencing (Table 1.). DNA libraries were prepared from various cervical cancer cell lines containing different HPV subtypes by using single index adapters (for MGI) and dual index adapters (for illumina) respectively, and then perform 5-Plex targeted capture sequencing.  It suggests that UDI strategy in illumina platform is a simple, feasible and robust way to handle the index misassignment.

Tab 1. The distribution of HPV calling from various cell lines after targeted capture sequencing

Solution

In order to settle the index misassignment, Nanodigmbio has issued MGI unique dual index (1-96) adapter module, known as the NadPrep Universal Adapter (MDI) Module (for MGI), in 2020. Recently, MDI adapters have been expanded to 768 varieties! With the help of the unique and corresponding index sequences at both ends of MDI adapters, MDI adapters can efficiently distinguish the reads with index misassignment from the reads with correct index sequences at both ends, making the subsequent analysis accurate without interference among samples. 

The patented design of unique dual index for MGI platforms, allowing 768 options!

With the popularization of DNBSEQ-T7 sequencer, known as “the world's best daily productivity sequencer”, the data output per lane has reached 1,440 Gb. Taking whole exome sequencing (such as Exome Plus Panel v2.0) as an example, at least 72 indexes are required for the WES at an average depth of about 250X(20 Gb/library). For the solid tumor detection application (such as NanOnco Plus Panel v2.0), a minimum of 288 indexes are required at an average depth of about 1,000X (5 Gb/library). Obviously, low-throughput index strategy cannot meet the needs of large number of libraries, especially in field of cancer research and genetic disease detection. MDI solution with 768 indexes  provide an easy way for various applications.

Table 2. Design rules of MDI

If the minimum balanced group is set to be 8, at least 8 continuous indexes have to be involved for multi-plex sequencing in one lane to achieve the best sequencing quality. The minimum balanced group of MDI designed by Nanodigmbio is only 4, indicating that the MDI adapters can be continuously used as long as ≥4 indexes. This design is user-friendly in most cases.

Uniform library yield and data output  

All the MDI have been verified by independent amplification to ensure uniform yield of the libraries (Fig. 2A). Meanwhile, when multi-plex sequencing was performed, data output are also similar(Fig. 2B).

Effectively improve data accuracy 

  To evaluate the overall accuracy and index misassignment of MDI adapters, multiple WGS libraries were prepared to simulate mixed sequencing in different cases. The results of split data from 12, 32 and 96 MDI libraries showed that nearly all the indexes in MDI adapters could be split correctly (>99.9%) (Fig. 3). MDI can greatly reduce the index misassignment, and promote the accuracy. Coupled with higher sequencing throughput, application of MDI makes clinical research more effective and low cost.

Low frequency mutation analysis  

Coupled with Nanodigmbio’s truncated adapters, MDI adapters can be applied in the targeted sequencing and low frequency mutation analysis of FFPE DNA or cell-free DNA from plasma/urine

Table 3. The application of MDI coupled with different adapters

Easy to use: 

In order to fulfill the convenience, MDI supports mixed sequencing together with the MGI UDB adapters. Similarly, MDI is also compatible with MGI single-index adapters. 

Reference: 

[1]https://www.illumina.com.cn/content/dam/illumina-marketing/documents/products/whitepapers/index-hopping-white-paper-770-2017-004.pdf?linkId=36607862

[2]Li Q, Zhao X, Zhang W, et al. Reliable multiplex sequencing with rare index mis-assignment on DNB-based NGS platform[J]. BMC genomics, 2019, 20(1): 1-13.

[3]MacConaill L E, Burns R T, Nag A, et al. Unique, dual-indexed sequencing adapters with UMIs effectively eliminate index cross-talk and significantly improve sensitivity of massively parallel sequencing[J]. BMC genomics, 2018, 19(1): 1-10.