Biomedical Engineering Reference
In-Depth Information
Infinium, GTYPE and Genotyping Console for Genechips, and GTGS for the MIP assay.
All programs are essentially similar, in that three clusters are automatically computed for
each SNP; the heterozygous AB and homozygous AA, or BB. The clusters are based upon
the allele-specific signal intensities. Genotyping errors and no-calls will hamper linkage and
association studies, and reliable SNP calls are essential for these applications. Therefore,
additional genotyping algorithms are becoming available to improve the quality of the
genotypes from SNP arrays. Examples of these methods are SNIPer [15] AccuTyping [16],
SNPchip [17] or RLMM [18].
3.2.3 Linkage and association analysis
The development of relatively inexpensive, high-throughput SNP arrays has transformed
genetic and genomic research. Their availability enables whole-genome association studies
for the unraveling of the complex genetics of disease. Success was reported for inflamma-
tory bowel disease [19] and type 2 diabetes mellitus [20] and recently for breast, prostate
and colorectal cancers [21-24]. An increasing number of reports of successful linkage
are appearing with a variety of phenotypes, such as Bardet-Biedl syndrome [25], neonatal
diabetes [26] or alcoholism [27]. In the study of colorectal cancer, arrays with over 10 000
SNPs were successfully used to identify a susceptibility locus on chromosome 3q21-q24
[28] and a locus on chromosome 10q23 that is associated with hereditary mixed polyposis
syndrome [29]. Linkage analysis using genotypes from SNP arrays can be performed using
freely available packages such as Mendel [30], Merlin [31] or Allegro [32]. In order to
facilitate the analysis and processing of genotypes of vast size, several freely available tools
have been developed; for example, Alohomora [33], SNPlink [34], CompareLinkage [35]
and Easylinkage [36]. The tools convert genotypes into the proper format for linkage pro-
grams, perform different levels of quality control and error removal, and graphically present
the linkage analysis data. SNPlink also automatically removes SNPs that are in high linkage
disequilibrium (LD) with nearby SNPs, since LD can falsely inflate linkage statistics.
3.2.4 Formalin-fixed, paraffin-embedded tissue
One increasingly important application is genotyping and genomic profiling of FFPE sam-
ples. Across the world, large collections of FFPE samples with clinical follow up exist in the
archives of pathology departments. Ideally, the DNA extracted from these tissues (see Pro-
tocols 3.1-3.3) might serve to perform linkage and association studies, or to generate tumor
profiles of LOH and genomic abnormalities. The availability of clinical follow up for these
samples will certainly strengthen the research. For linkage analysis, the use of FFPE tissue
for genotyping also allows incorporation into analyzes of individuals for which leukocyte
DNA is unavailable. A major drawback of the use of DNA from FFPE samples is DNA
quality (see Protocol 3.4). As a consequence of formalin fixation of the tissue, extracted
DNAs show varying levels of degradation. The degree of degradation depends largely on
the length and the method of fixation, and the age of the specimen. The implication for SNP
arrays is that not all the methods are suitable for use (or will not perform consistently) with
FFPE DNAs. Whereas the high-density Genechip and Infinium arrays are designed for use
with high-quality genomic DNA, both the Goldengate and the MIP assay can be used for
genotyping and for detection of LOH and copy number changes in FFPE tissue [37-39].
