Premium
Copy number estimation algorithms and fluorescence in situ hybridization to describe copy number alterations in human tumors
Author(s) -
Suzuki Masaya,
Nagura Kiyoko,
Igarashi Hisaki,
Tao Hong,
Midorikawa Yutaka,
Kitayama Yasuhiko,
Sugimura Haruhiko
Publication year - 2009
Publication title -
pathology international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.73
H-Index - 74
eISSN - 1440-1827
pISSN - 1320-5463
DOI - 10.1111/j.1440-1827.2009.02354.x
Subject(s) - copy number variation , copy number analysis , fluorescence in situ hybridization , algorithm , genotyping , dna microarray , comparative genomic hybridization , biology , computational biology , computer science , replicate , genetics , mathematics , genotype , genome , statistics , gene , chromosome , gene expression
The platforms of high‐resolution genetic analysis of human tumors have become popular, and several copy number estimation algorithms have been applied to the data generated by single‐nucleotide polymorphism microarrays. Although comparisons have been made between several different platforms or methodologies, there has never been a robust comparison of different copy number estimation algorithms, and the validity of the estimations in comparison with multiple fluorescence in situ hybridization (FISH) data in tumors has rarely been addressed. In the present study the dataset that the Affymetrix 250K Nsp array generated in two cancer cases was used to compare the two widely used algorithms for estimating copy number alterations (CNA): the genotyping microarray‐based copy number variation (CNV) analysis (GEMCA) algorithm and the copy number analyzer for Affymetrix Genechip mapping (CNAG) algorithm. Considerable differences were noticed between the estimations by these two algorithms, because of the difference in the formula used to calculate the threshold values. Both algorithms yielded highly consistent data with the FISH results, but CNAG was more stringent for detecting loss. There were areas in which both algorithms provided gains, but FISH showed no change. It will be interesting to pursue the reasons for these remaining discrepancies.