Editorial Article
Wednesday August 25, 2010
by Caitlin Smith
Comparative genomic hybridization (CGH) is a long name for a process that has happily become shortened and vastly more productive in the past few years. It uses microarrays of oligonucleotide probes to examine things such as copy number variation (CNV), amplifications, and deletions in genetic codes. Today, array-based CGH (aCGH) can do this more efficiently and with greater resolution than older techniques such as karyotyping, or fluorescence in situ hybridization. “Accurate determination of copy number aberrations is important for understanding developmental disorders, oncogenesis, and normal genetic variation,” says Anniek De Witte, CGH product manager at Agilent Technologies. “The main benefit of using arrays is that they are more powerful, because the technology is high-resolution and can detect smaller duplications and deletions, which were previously missed by karyotyping or other traditional techniques. As a result, the International Standard Cytogenomic Array (ISCA) Consortium now recommends that microarrays should replace G-banded karyotypes as the first-tier technology to study individuals with unexplained developmental delays and intellectual disability, autism spectrum disorders, or multiple congenital anomalies.” Yet aCGH is still evolving rapidly, and more is in store for researchers to learn.
Why use aCGH?
One reason to use aCGH is the high-quality resolution. Ruth Burton, product manager for clinical solutions at Oxford Gene Technology (OGT) agrees that “the primary advantage of aCGH is the substantial increase in overall resolution and diagnostic yield.” In addition to a range of aCGH products, OGT’s Genefficiency™ Service combines complete design, project management, and data analysis. “We were entrusted with the world's largest copy number variation study where over 2 billion high-quality data points were generated from >20,000 samples in just 20 weeks for the Wellcome Trust Case Control Consortium,” says Burton. Research scientists are currently using OGT’s CytoSure™ product range to identify genetic aberrations causing a range of developmental disorders (e.g., DiGeorge and Williams-Beuren). “Our latest CytoSure ISCA arrays, designed in collaboration with the ISCA Consortium, provide standardized evidence-based formats focusing on disease and syndrome-associated genome regions, in addition to offering whole-genome ‘backbone’ coverage,” says Burton.
Another reason to use aCGH is that it is becoming increasingly more affordable, and more customizable. Agilent’s SurePrint G3 Custom CGH microarrays let you design your own custom microarrays with their eArray online array creation application tool, which lets you choose from their database of over 24 million predesigned, in silico-validated CGH probes. “The Agilent two-color CGH microarray platform provides superior copy number change detection,” says De Witte, “thanks to Agilent’s high-fidelity 60-mer probes synthesized through inkjet printing technology. Customization will continue to be a key differentiator.”
Uncovering new information about CNV, perhaps in relation to disease, is another reason to use aCGH today. “Microarray resolution has been increasing over time; oligo arrays keep getting denser,” says Lance Brown, marketing manager at Roche. “So in many experiments performed, say, three years ago, there were many CNVs present that were never found – that they weren’t able to detect with that resolution.” Arrays with progressively better resolution are generating progressively more data as they find the smaller CNVs that were missed the first time. “It’s helping to explain a lot of the questions that the disease researchers have had, but could not find,” says Brown.
Roche’s current aCGH technology is a 2.1 million feature microarray, which on average allows one to detect copy changes as small as 5000 bases. They also offer their new NimbleGen MS 200 microarray scanner, with 2 micron resolution, for good feature detection. In addition, Roche has a new CNV-focused array (also 2.1 million features), designed with probes concentrated in regions containing known CNVs (more than 41,000 CNVs). “A lot of CNV research in the past, and continuing today, has been to discover CNVs,” says Brown. “But due to the high density that we have, we were able to put a pretty dense backbone across the whole genome as well, for the discovery of unknown CNVs. So it’s a nice combined design: it allows you to cover all the known CNVs that have been well-documented, but still discover new ones.”
Making aCGH more powerful
Despite its mind-boggling numbers, aCGH will yet become more powerful. “Although aCGH offers many advantages over SNP-based approaches for detecting CNV changes – including superior sensitivity, more comprehensive genome coverage, and more accurate quantification of CNV changes – SNP platforms allow detection of uniparental disomy,” notes Burton. “Uniparental disomy, where two copies of a full or part chromosome are received from one parent, is linked to a number of diseases, including Prader-Willi and Angelman syndrome. OGT will shortly be announcing an array that combines the benefits of both aCGH and SNP technologies to enable superior aberration detection combined with reliable uniparental disomy identification.”
Array CGH may also become a powerful tool in fertility research. “There is also growing evidence to suggest that aCGH analysis of pre-implantation embryos could offer significant improvements over traditional techniques in detecting genetic abnormalities prior to implantation, thereby increasing pregnancy success rates,” says Burton. “Preimplantation genetic screening is often offered to couples with known genetic disorders or with difficulties conceiving. The challenge with this technique is the accurate amplification of small amounts of DNA from a single, early-stage embryo cell.”
Roche plans to make aCGH more powerful in the near future by doubling the density of their aCGH microarrays to 4.2 million features. “We’ll be able to detect more routinely the smaller CNVs,” says Brown. “That opens up new information that wasn’t found in the past.” Their yet-to-be-released 4.2 million feature arrays will have a probe about every 600 bases. “You need about 5 probes in a row to make a call that maybe there is an amplification or a deletion present,” says Brown. “So taking that as a rule of thumb, today we can detect about 5000 base copy number changes routinely. When we move to a 4.2 million feature array, we’ll be able to move down to half that size to about 2000-3000 bases for copy number changes.”
The challenges of so much data
Generating so much data then generates a new challenge – what to do with it. Bioinformatics have never been so important. “Now, more than ever with aCGH, researchers really need to think about the bioinformatics needs up front,” says Brown, “and create a study design, and the support they need to generate useable data, to get the data into a format that’s useable for publication.” Burton agrees that “due to the sheer amount of data available from aCGH, the biggest challenge is often the interpretation of results.” OGT's CytoSure Interpret Software is a user-friendly package for analyzing aCGH data.
De Witte also agrees that data interpretation is one of the next big challenges for aCGH scientists. “Although labs using aCGH are very confident in the quality and the reproducibility of the microarray data, users are sometimes challenged by the interpretation of the array data. More data is constantly being added to databases like the ISCA database, the Database of Genomic Variants, and DECIPHER (Sanger Institute, UK). Using these databases, consortia like ISCA will develop recommendations for the interpretation and reporting of copy number changes.”
De Witte believes that CGH microarrays are here to stay. “Next generation sequencing is a very powerful tool but the cost, turnaround time, and complexity of the data remain big challenges that need to be overcome before next-generation sequencing can even begin to achieve the throughput that microarrays can today.”
The image at the top of this page is from Agilent's Comparative Genomic Hybridization webpage. The figure is referenced to be from, "The pitfalls of platform comparison: DNA copy number array technologies assessed," BMC Genomics 2009, 10:588.