Accurate Jitter Computation in CNA Breakpoints Using Hybrid Confidence Masks With Applications to SNP Array Probing

Abstract

Chromosomal structural changes known as copy number alterations–aberrations (CNAs) result in gains or losses in copies of deoxyribonucleic acid sections, which are typically associated with different types of cancer. An intensive noise inherent to modern technologies of CNAs probing often causes inconsistency between the estimates provided by different methods. Therefore, testing estimates by the confidence masks is recommended to guarantee an existence of genomic changes within certain regions. In known masks, jitter in the CNA’s breakpoints is expected to be distributed with the skew Laplace law, which is sufficiently accurate when the segmental signal-to-noise ratio (SNR) exceeds unity. In this paper, we extend the confidence masks to low and very low SNRs often observed in subtle chromosomal changes. The modified masks employ several proposed approximations of the segmental noise variance as a function of the departure step from the candidate breakpoint. Because approximations are accurate in jitter computation only for specified SNR regions, we suggest using hybrid masks to achieve the maximum available accuracy. Confidence masks are tested experimentally by genome CNA profile data obtained using the single nucleotide polymorphism array.