1. Introduction
The progressive accumulation of genetic changes plays an essential role in the tumorigenesis and evolution of human cancers. The genetic changes commonly seen in human cancers include chromosomal translocations, amplifications, allelic loss, loss of heterozygosity, deletions, mutations, and epigenetic changes/DNA methylation affecting oncogenes and tumor suppressor genes [1]. The resolution of genetic alterations identified in clinical specimens has been pushed to the single nucleotide level over the decades with advancements in genetic technologies. Conventional cytogenetic analysis with G-banding karyotyping, a routine clinical analysis in cytogenetic labs, allows differentiation of approximately 400–500 bands per haploid genome [2]. At this level of resolution, chromosomal change over 10 Mb can be detected. Fluorescence in situ hybridization (FISH) offers high sensitivity and specificity of detecting genetic abnormalities such as translocations, aneuploidy, deletions, inversions, or amplifications by using DNA probes targeted to known DNA sequences [3]. FISH can identify genetic changes at a resolution up to a few kilobases (kb), but is not suited for identification of unknown genetic changes or global chromosomal abnormalities. Array-based comparative genomic hybridization (aCGH) developed in the early 1990s offers efficient high-throughput analysis of the entire genome for identification of copy number variations/aberrations (CNVs/CNAs) that are usually not detectable by conventional karyotyping or targeted FISH studies, and has an improved resolution down to 100 kb [4,5,6,7]. Single nucleotide polymorphism (SNP) arrays, manufactured by Affymetrix and Illumina, were initially designed for high-throughput SNP genotyping, but were quickly applied to cancer genomics [8,9,10,11]. In contrast to aCGH, SNP arrays are able to detect both CNVs/CNAs and loss of heterozygosity (LOH) or copy-neutral LOH/uniparental disomy (UPD), which are frequently involved in the development of cancers. With the advance in technology and marked improvements in resolution, the new SNP array offers over 90% coverage of known copy number variants by using more than 946,000 probes and an average inter-marker distance of 680 base pairs. This high level of resolution of cytogenetic changes has only recently been surpassed by next generation sequencing (NGS) technology developed in the last decade [12,13]. Ever since the invention of SNP arrays, they have been extensively applied to various hematologic malignancies. While currently there are no clinical guidelines on the use of SNP array in hematopoietic malignancies, SNP array will certainly be useful in difficult cases, especially in myelodysplastic syndrome (MDS) diagnosis, when other methodologies fail to identify cytogenetic abnormalities. A proposed flow chart for the application of SNP array in hematopoietic malignancies is presented in Figure 1. In this review, we summarize the important findings of chromosomal changes in hematopoietic malignancies identified by SNP array analysis.
Figure 1  Proposed application of SNP array in hematopoietic malignancies.

2