Abstract 3648: Resolving copy number variations from existing exome SNP data without the need for sophisticated software

Abstract

Abstract Several tools exists to evaluate copy-number variation (CNV) from sequencing read data. Here, the aim is to investigate whether larger CNVs in leukemia can be extracted directly from already called variants - without specialized algorithms. Methods: 12 cases of leukemia, with 6 paired and 6 unpaired, were included (Table 1). Following exome sequencing and reference alignment variants were called with GATK to produce recalibrated sets in variant call format (VCF). Passed single nucleotide polymorphisms (SNPs) were limited to depth of coverage > 29x. Each chromosome were visualized as frequency scatter plot and smoothened histograms of frequency distribution in order to inspect frequency bands and number of peaks in histograms. Kolmogorov-Smirnov test was employed to compare distributions between paired samples. Results: None of the AML cases with normal karyotype displayed CNVs and no significant changes in frequency distributions were detected. In the relapse of paired T-ALL a Chr 4 trisomy and a 33×10−6 bp copy-number neutral loss of heterozygosity (CN-LOH) on Chr 9 were detected. Collectively, 7 aberrations confirmed standard cytogenetics, 8 new aberrations were added to the cytogenetical profile and 5 could not be detected as shown in Table 1. The aberrations not detected were, as expected, balanced translocations (3) and microdeletions (2). Discussion: No previously found large aberrations escaped detection, when disregarding translocations. In addition, other CNVs not detected by conventional cytogenetics were found by SNP analysis. Moreover, the added information is in agreement with the literature: 5% of T-ALL patients display del(11)(p12p13) and +12 is one of the most common cytogenetic changes of CLL, e.g. Thus, we suggest that CNV profiling can be initiated by directly evaluating frequencies from existing variant calls. This may ease implementation of “closed toolboxes”, enable analysis of single unpaired samples and function as a direct complement to array-CGH or FISH to detect CN-LOH. Table 1SamplesPairedCytogeneticsCNV detectedNK-AML5YesNormal karyotypeNoneT-ALL relapse1Yes4+, 3p26.2 and Xp22.2 micro-deletions+4, partial del(9p)AML M01YesNormal karyotypePartial del(11p)AML M31Not(15;17)NoneCML1Not(9;22)NoneT-ALLNodel(5) t(7;14)Partial del(5q), +4q, +9p, del(11p), del(12q)CLL1NoNot analyzed+12AML M01NoNormal karyotypePartial +11 Citation Format: Marcus C. Hansen, Laura L. Herborg, Anne S. Roug, Sadudee Chotirat, Chirayu U. Auewarakul, Eigil Kjeldsen, Peter Hokland. Resolving copy number variations from existing exome SNP data without the need for sophisticated software. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3648.