Details

Autor(en) / Beteiligte

• Akimov, Yevhen
• Eldfors, Samuli
• Malani, Disha
• Aittokallio, Tero
• Heckman, Caroline

Titel

Abstract 3277: Identification of internal tandem duplication within the FLT3 gene from AML patient next-generation sequence data

Ist Teil von

• Cancer research (Chicago, Ill.), 2018-07, Vol.78 (13_Supplement), p.3277-3277

Erscheinungsjahr

2018

Quelle

EZB Electronic Journals Library

Beschreibungen/Notizen

• Abstract Introduction The presence of the Fms Related Tyrosine Kinase 3 (FLT3) internal tandem duplication (FLT3-ITD) is associated with a poor prognosis and an increased rate of relapse in acute myeloid leukemia (AML). Identification of FLT3-ITD from next-generation sequencing (NGS) data is currently impeded by lack of reliable algorithms to identify large insertions. Therefore, despite the wide adoption of NGS in recent years, FLT3-ITD mutations are still being detected with PCR/electrophoresis assays which are expensive and time consuming. Algorithms that reliably identify FLT3-ITDs from NGS data are urgently required. Methods We developed a computational algorithm for FLT3-ITD calling from next-generation sequencing data. The method is based on iterative realignment of unaligned sequence read fragments. The algorithm uses sequence reads mapped to FLT3 exon 14 and unmapped sequence reads sharing similarity (at least 25bp aligned with the Smith-Waterman algorithm) with FLT3 exon 14. The procedure has two steps: (1) Smith-Waterman local pairwise alignment of each read to the reference sequence, and (2) trimming unaligned parts of each read. If a read has an unaligned part(s) identified at step 2, then these are returned to the step 1, and the whole procedure is repeated. The procedure is iterated until each read is fully-aligned to the reference sequence. A sample is considered to be ITD negative if all reads are fully-aligned after the first iteration. Otherwise, the algorithm returns fragmentized reads aligned to the reference sequence. This output can be used to discriminate between ITD positive and negative samples based on the pattern produced by aligned fragments of the reads. Reads coming from FLT3-ITD have an intrinsic property to cluster at the boundaries of the duplicated region; therefore, samples with reads ends clustering at a shared genomic position are considered ITD positive. We calculate the p-value of reads ends clustering at the same genomic position using generalized birthday-problem theory. Results We analyzed a total of 273 AML samples from 125 patients. The algorithm assigned 221 samples (~81%) as FLT3-ITD negative, forty-seven (~17%) samples as FLT3-ITD positive and 5 samples (~1.8%) were assigned as undefined (with a p-value between 0.005 and 0.1). We visually inspected the patterns produced by realigned fragments to assign a FLT3-ITD status for each of the undefined samples. We compared these results with FLT3-ITD status determined by a clinical PCR/electrophoresis assay. Our algorithm was able to recall all the 13 samples where ITD was detected by the clinical assay. Notably, FLT3-ITD mutation in two samples were detected with our method but not by the clinical assay. Conclusions 1. Our novel algorithm enables accurate identification of FLT3-ITD mutations from NGS sequence data. 2. The method was able to detect FLT3-ITD positive samples that were missed by the clinical assay. Citation Format: Yevhen Akimov, Samuli Eldfors, Disha Malani, Tero Aittokallio, Caroline Heckman. Identification of internal tandem duplication within the FLT3 gene from AML patient next-generation sequence data [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 3277.