Beschreibung:
<jats:title>Abstract</jats:title>
<jats:p>Due to recent scientific advances, there is a growing demand for new indices with high fidelity to eliminate errors associated with certain index combinations and sequencing chemistries. These errors are due to insufficient edit distances (the minimum number of changes required to transform one index sequence to the other) between index sequences or index hopping. Recent publications have highlighted how sequencing reads are misassigned due to "index hopping" on Illumina patterned flow cells, such as the recently launched Illumina Novaseq which can generate billions of reads (&gt;100 exomes) in a single run. This misassignment canlead to false positives in ultra-sensitive assays where low frequency variants or nucleic acid species are monitored. We have observed misassignment due to insufficient edit distance among Illumina TruSeq HT i7 indices at a frequency up to 1.5%. Therefore, we developed 968nt i7 indices that can be paired with the existing TruSeq HT i5 indices to achieve 768 high throughput dual combinations, which were validated using a novel method on both Illumina's 2- and 4-channel technologies as both single and dual indices. This method involved thepreparation of 96 libraries with unique, non-overlapping inserts to facilitate tracking of index misassignment. This allowed us to assess not only which index has misassigned library molecules, but also pinpoint the origin and rate of misassignment within a single run. With our 96 i7 indices, misassignment was observed at rates &lt;0.1%. For even higher fidelity de-multiplexing, we have paired our 96 indices in a non-tandem manner for single use in both the i5 and i7 positions, known as Unique Dual Indices (UDIs). The use of UDIs further eliminates index read errors that misassign reads, enabling increased confidence in calling low frequency variants. UDIs can also eliminate PCR-induced chimerisms, which can significantly improve data from a variety of assays. We are validating 96 new indices as UDIs for avoidance of index hopping and for eliminating PCR-induced chimerism during multiplexed library amplification.</jats:p>
<jats:p>Citation Format: Jonathan C. Irish, Jordan RoseFigura, Sukhinder K. Sandhu, Bita Carrion, Laurie Kurihara, Vladimir Makarov. Improved sample indexing for high fidelity demultiplexing to increase confidence in low frequency variant calling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1430.</jats:p>