An optimization framework for unsupervised identification of rare copy number variation from SNP array data

Abstract

A highly sensitive and configurable method for calling copy number variants from SNP array data is presented that can identify even rare CNVs

Genome-wide detection of copy-number variations in local cattle breeds

“…Comparative analyses of the algorithms used for CNV identification have been published (Winchester et al 2009;Xu et al 2013). As reported by several authors (Marenne et al 2009;Yavaş et al 2009;Seroussi et al 2010;Ma et al 2017), PennCNV is the most reliable and accurate algorithm in detecting CNVs from Illumina BeadChip data. PennCNV is a tool that incorporates multiple sources of information, such as the signal intensity at each SNP marker, the distance between neighbouring SNPs, and the BAF, and integrates a computational approach by fitting regression models of the GC content to avoid 'genomic waves'.…”

Contrasting analytical and data-driven frameworks for radiogenomic modeling of normal tissue toxicities in prostate cancer

“…The objective of CNV analysis is to identify the chromosomal regions at which the number of copies of a gene deviates from two. These could be gains (CNV >2) or losses (CNV <2) [18]. The role of CNVs has only recently been implicated in inflammation-related diseases [18] and has not yet been investigated in RITs.…”

Optimization Algorithms for Identification and Genotyping of Copy Number Polymorphisms in Human Populations

“…We use three different algorithms, ÇOKGEN [7], PennCNV [6] and Birdseye [4] to detect the initial set of CNVs that serve as input to POLYGON.…”

“…To date, several methods have been proposed for inferring CNVs from SNP array data [3][4][5][6]. In a recent study [7], we have formulated CNV identification as an optimization problem with an explicitly designed objective function that is characterized by several adjustable parameters. Our method, ÇOKGEN, efficiently identifies CNVs using a variant of the well-known simulated annealing heuristic.…”

“…The proposed method, POLYGON, takes as input the copy number variants identified by a single-sample CNV identification algorithm (e.g., ÇOKGEN [7], PennCNV [6], Birdseye [4]) and implements a computational framework to (i) identify CNVs in different samples that might correspond to the same variant in the population (candidate CNPs), (ii) adjust the boundaries of these candidate CNPs by drawing strength from raw copy number data from multiple samples, and (iii) determine copy number genotypes in the study. The key ingredient of this computational framework is an explicitly formulated objective function that takes into account several criteria, which are carefully designed to quantify the desirability of a CNP genotype with respect to various biological insights and experimental considerations.…”