Comparative analysis of algorithms for identifying amplifications and deletions in array CGH data - PubMed (original) (raw)

Comparative Study

Comparative analysis of algorithms for identifying amplifications and deletions in array CGH data

Weil R Lai et al. Bioinformatics. 2005.

Abstract

Motivation: Array Comparative Genomic Hybridization (CGH) can reveal chromosomal aberrations in the genomic DNA. These amplifications and deletions at the DNA level are important in the pathogenesis of cancer and other diseases. While a large number of approaches have been proposed for analyzing the large array CGH datasets, the relative merits of these methods in practice are not clear.

Results: We compare 11 different algorithms for analyzing array CGH data. These include both segment detection methods and smoothing methods, based on diverse techniques such as mixture models, Hidden Markov Models, maximum likelihood, regression, wavelets and genetic algorithms. We compute the Receiver Operating Characteristic (ROC) curves using simulated data to quantify sensitivity and specificity for various levels of signal-to-noise ratio and different sizes of abnormalities. We also characterize their performance on chromosomal regions of interest in a real dataset obtained from patients with Glioblastoma Multiforme. While comparisons of this type are difficult due to possibly sub-optimal choice of parameters in the methods, they nevertheless reveal general characteristics that are helpful to the biological investigator.

PubMed Disclaimer

Figures

Fig. 1

Fig. 1

Array-CGH algorithms on simulated aberrations of increasing width. Illustrated here as an example are the signal profiles consisting of five aberrations of 2, 5, 10, 20, and 40 probes long with an amplitude of 1. Gaussian noise N(0, .252) was added onto the signal profile to generate the simulated data. Default settings for the algorithms were used when available; otherwise, appropriate parameters were selected or computed based on the program documentation and related papers.

Fig. 2

Fig. 2

Receiver operating characteristic (ROC) curves for array CGH algorithms measured at different aberration widths and signal-to-noise ratios (SNR). The _x_-axis is the false positive rate and the _y_-axis is the true positive rate. Red is CGHseg (Picard et al., 2005), orange is quantreg (Eilers and de Menezes, 2005), dark yellow is CLAC (Wang et al., 2005), green is GLAD (Hupe et al., 2004), blue is CBS (Olshen et al., 2004), violet is HMM (Fridlyand et al., 2004), salmon is wavelet (Hsu et al., 2005), black is lowess, light green is ChARM (Myers et al., 2004), brown is GA (Jong et al., 2003), and cyan is ACE (Lingjaerde et al., 2005). The curves were generated by measuring the true and false positive rates on simulated data at different threshold levels.

Fig. 3

Fig. 3

Array-CGH profile of chromosome 13 in a Glioblastoma Multiforme sample (GBM31). This chromosome has a partial loss of low magnitude. Most algorithms in the study detect the loss. In particular, CGHseg, GLAD, CBS, and GA clearly identify the region.

Fig. 4

Fig. 4

Array-CGH profile of the three amplifications around EGFR in GBM29. CGHseg, quantreg, GLAD, wavelet, and GA detects all three amplifications. CLAC, CBS, Lowess, and ACE detect the first two amplifications as one larger region. ChARM detects the amplification as one large region of gain, while HMM does not detect any.

Similar articles

Cited by

References

    1. Autio R, Hautaniemi S, Kauraniemi P, Yli-Harja O, Astola J, Wolf M, Kallioniemi A. CGH-Plotter: MATLAB toolbox for CGH-data analysis. Bioinformatics. 2003;19:1714–1715. - PubMed
    1. Beheshti B, Braude I, Marrano P, Thorner P, Zielenska M, Squire JA. Chromosomal localization of DNA amplifications in neuroblastoma tumors using cDNA microarray comparative genomic hybridization. Neoplasia. 2003;5:53–62. - PMC - PubMed
    1. Bredel M, Bredel C, Juric D, Harsh GR, Vogel H, Recht LD, Sikic BI. High-resolution genome-wide mapping of genetic alterations in human glial brain tumors. Cancer Res. 2005;65:4088–4096. - PubMed
    1. Brennan C, Zhang Y, Leo C, Feng B, Cauwels C, Aguirre AJ, Kim M, Protopopov A, Chin L. High-resolution global profiling of genomic alterations with long oligonucleotide microarray. Cancer Res. 2004;64:4744–4748. - PubMed
    1. Cleveland WS. Robust locally weighted regression and smoothing scatterplots. J. Amer. Statist. Assoc. 1979;74:829–836.

Publication types

MeSH terms

LinkOut - more resources