New Array Can Give Detailed Genetic Profile of Any Mouse

September 8, 2009

High-Resolution Genotyping for the Mouse

Containing approximately 7,000,000 probes on a silicon wafer about a half inch square, the Mouse Diversity Genotyping Array is the most advanced high-density mouse genome-wide profiling array available. It can simultaneously assay approximately 625,000 single nucleotide polymorphisms (SNPs) and over 900,000 invariant genomic probes (IGPs) in the mouse genome. It can reveal the genetic profile of virtually any mouse in the world — existing, novel, domestic, wild, or a complex mixture of these. In early 2009, we began offering services to researchers wishing to make use of the array. In early August, the rationale and methods for developing the array were published (Yang et al. 2009). This article briefly discusses why and how the array was constructed, its utility in mouse-based biomedical research, and the services JAX offers with the array.

chip and array

Background: Raw image data obtained from a laser scan of a portion of the Mouse Diversity Genotyping Array. The entire array is tiled (populated) with more than seven million 25-base pair oligonucleotide probes, which can assay nearly 625,000 SNPs and more than 900,000 invariant genomic probes (IGPs). The probes bind to labeled DNA from a target sample. Analysis software is used to quantify the brightness of each fluorescing DNA-probe complex (seen here as spots) on the gridded image. High intensity spots indicate high affinity between the probe and target DNA sequences and are used to decode the genotypes of individual SNPs.

To the right: superimposed on the gridded image, is an Affymetrix chip. The silicon wafer is housed inside, under a transparent window. The DNA sample is inserted through a hole on the back of the chip.

Rationale

In spite of all the recent advances in mouse genetics, mouse geneticists have not been able to identify genes as successfully as hoped for in genome-wide association studies of outbred mouse populations. They needed a high density/high diversity genotyping SNP array, similar to the human HapMap array, which contains probes for a large quantity of SNPs ascertained from large and widely spread human populations. Hence, it can be used to produce a high-resolution genetic profile of virtually anyone in the world. In contrast, available mouse arrays contained many fewer SNP probes, and these probes did not represent SNPs from wild-derived and wild-caught mice. Hence, their utility for genotyping wild-caught or wild-derived mice, revealing representative allele frequencies, producing highly accurate mouse phylogenies, performing genome-wide association studies, and genotyping Collaborative Cross and Diversity Outcross Mice (both of which contain significant genomic components from wild-caught and wild-derived mouse strains) were limited. A more useful mouse array had to be capable of assaying significantly more of the estimated 45,000,000 SNPs in the mouse genome, including/especially those in wild mice. A team (hereafter called the "team") led by Professor Gary Churchill of the Jackson Laboratory and Dr. Fernando Pardo-Manuel de Villena, University of North Carolina, both of The Center for Genome Dynamics, set out to construct such an array.

churchill fernando

The Mouse Diversity Genotyping Array was designed by Professor Gary Churchill Ph.D. (left), The Jackson Laboratory, and Dr. Fernando Pardo-Manuel de Villena (right), University of North Carolina, both of The Center for Genome Dynamics, in cooperation with Affymetrix, and was funded by the National Institute of General Medical Sciences.

Selecting the SNP Probes

To choose which SNPs to assay from the estimated 45,000,000 in the mouse genome, the team scanned the genome of the 15 strains (many of them wild-derived strains) sequenced by the NIEHS and Perlegen and the vast amount of mouse SNP data — including those from wild mice — available online. Ultimately, they chose to "tile" (populate) the array with probes for 623,124 highly polymorphic SNPs spaced approximately one every 4.3kb across all parts of the mouse genome. The three or four major subspecies of mice distributed throughout the world, the major branches of the mouse phylogenetic tree, and wild mice are all very well represented.

Balance Between High Density and Redundancy

There is only so much room on an array chip. Although the team wanted an array that was capable of high resolution (probes for many SNPs closely spaced throughout the genome), they also wanted an array that provided enough redundancy (several randomly placed probes for each SNP) to yield robust statistical power and to compensate for array scratches, probe failures, or anomalous reading. Too much resolution would mean too little redundancy and vice versa. To strike a balance, they decided to tile an average of eight probes for each SNP, four for each of two alleles at each SNP: two of the four probes for each allele were for the SNP on one DNA strand, and two were for the same SNP on the complimentary DNA strand. (Assaying both DNA strands is turning out to be more beneficial than merely providing redundancy: it is yielding unexpected information about unusual/little-understood features of the mouse genome—the subject of a future article. Given that the array has an average of eight probes per SNP, the total number of SNP probes on the array is approximately 5,000,000.

More Array Real Estate Becomes Available

In the process of constructing the array, Affymetrix's highest-density array format (the same format used for the HapMap array) became available, allowing the team to add nearly two million more probes to the array. So, the team added 916,269 invariant genomic probes (IGPs) to the array (actually, 1,832,538 — one for each of the two complimentary DNA strands). These were probes for regions outside of SNPs: three probes for every exon—so the array could be used to assay gene expression; probes in ultra conserved elements (UCEs)—regions that theoretically hadn't changed since the Devonian Period, (spanning 416 to 359.2 million years ago, when fish evolved); and several thousand probes in segmental duplications—genomic regions where DNA is duplicated and where copy number variations (CNVs) are expected to be most common. These probes will greatly increase the utility of the array for understanding little-explored genomic regions.

The Array's Quality

Affymetrix had learned a great deal from their experience developing other arrays, including human, dog, and other mouse arrays. This experience is reflected in the excellent results so far obtained from the Mouse Diversity Genotyping Array. Statistical analyses and various other criteria indicate that, as long as the DNA is very carefully prepared, the array performs extremely well. As Professor Churchill states, "The array has been used over 2,000 times with remarkable success". For example the team is currently genotyping some 70 mouse strains that, as Professor Churchill states "come from the bushes" —Kazakhstan, Southeast Asia, Arizona, and many other places around the world—and the array is generating excellent genetic profiles of these mice. Indeed, even what were initially thought to be probe failures yielded unexpected information. The 25-base pair oligonucleotide probes were designed so that the middle base would bind to one of two SNP alleles in a DNA sample. However, DNA in some samples bound to probe bases flanking the middle base, indicating that the flanking bases were binding to previously uncharacterized polymorphisms.

Watch a video of Professor Churchill answer questions about the array.

Uses of the Array

Many customers have contacted JAX to ask what the Mouse Diversity Genotyping Array's capabilities are. They are many, including the following:

  • Performing high resolution mapping and genetic analyses
  • Analyzing CNVs
  • Carrying out association and quantitative trait loci (QTL) analyses
  • Studying population genetics
  • Performing genetic quality control
  • Characterizing or comparing mouse DNA samples (e.g., from tumor tissues, cell lines, substrains, or sister strains)

The array will be particularly useful for genotyping wild mice and the Collaborative Cross and Diversity Outcross mouse populations, both of which contain a large number of recombinations and significant genetic contributions from wild-derived mice. The array is being used to better characterize recombinant inbred and consomic strains. It promises to yield many new insights into ultraconserved regions, gene methylation, and gene-environment interactions. It will significantly refine and increase the power of complex trait analysis.

JAX® Services currently offers Diversity Array services, including DNA preparation, hybridization, and basic or custom SNP data analysis services. To find out more about the services we offer, visit the array website, or contact us at 1-800-422-6423, 1-207-288-5845, or jaxservices@jax.org.

Reference

(Authors in bold are Jackson Laboratory scientists.)

Yang H, Ding Y, Hutchins LN, Szatkiewicz J, Bell TA, Paigen BJ, Graber JH, PardoManuel de Villena F, Churchill GA. 2009. A customized and versatile high density genotyping array for the mouse. Nat Meth Aug 9. [Epub ahead of print].