Affiliations: Department of Computer Science, University of Texas at
San Antonio, TX 78249, USA | Live Search, Microsoft Corporation, One Microsoft way,
Redmond, WA 98052, USA
Abstract: DNA copy number aberrations along the genome are vital markers for
studying pathogenesis of various diseases including cancers. Array-based
Comparative Genome Hybridization (aCGH), which is a high-throughput cytogenetic
method, helps in identifying genome-wide copy number aberrations, both gains
and losses. Here, we propose a computational technique to analyze aCGH data and
to identify potential DNA copy number alterations along the genome. Our
technique detects the possible breakpoints by comparing contiguous probe log
ratios, reports the aberrant segments and handles outliers to minimize false
discovery rate. Empirically, we tested our algorithm on both prokaryotic (Brucella ovis)
and eukaryotic (glioblastoma and colorectal cancer datasets
from human) genomes. Our findings complement previous studies; our performance
is competitive, sometimes superior, against other popular methods.
Keywords: aCGH, chromosomal aberration, DNA copy number
