Using arrays of DNA to interrogate gene expression was first described in 1987 [1]. These gene arrays were realized as cDNAs spotted onto filter paper. Today's microarray technology was first used for gene expression profiling in 1995. Since then, the use of microarrays skyrocketed and the technology quickly became an integral part of many research pipelines and efforts. Microarrays are a powerful tool for applications such as transcriptome analysis, profiling DNA-protein interactions and genetic variation analysis, both large (CNV) and small (SNP). Microarrays provide researchers with a massively parallel tool for analyzing vast amounts of genetic information in a single experiment. But this power is not without some inherent limitations: because microarrays rely on hybridization, some knowledge of the sequences being investigated is required; highly similar sequences can result in cross-hybridization; and analog signal problems can exist that make it difficult to confidently interpret results [2]. Nevertheless, analysis techniques were created, experiments designed and results interpreted; microarrays made their impact and contribution to genomics research for over two decades.

Since 2004, several next-generation high-throughput sequencing platforms (NGS) have been created that offer massively parallel short-read DNA sequencing to interrogate genomes. NGS technologies currently have their own limitations, mainly with data analysis and storage, but they do offer solutions to the microarray limitations: no knowledge of the genome sequence is required; NGS is direct sequencing and not hybridization so there is no worry of cross-hybridization occurring; and quantification of the signal is digital, based on counting sequence tags [3]. But of course one of the biggest advantages of NGS is the sheer volume of data produced from a single run, providing high coverage and resolution. With all this power, NGS is making its way into areas traditionally dominated by microarrays.

Recently, Illumina announced the release of a new high-throughput sequencer, the HiSeq™ 2000 [4]. While not a ground-breaking release for the technology, it does significantly lower the cost-per-base of a run, moving NGS to a more affordable level and opening it up to smaller researchers. In light of this new release, a question has been frequently asked: Is this the end of microarrays? The answer seems to be, it depends on what you want to do. Even these lower costs for using NGS are still prohibitive for many researchers and microarrays are widely used and familiar, with many bioinformatic tools and pipelines already created. For some experiments and smaller genomes, NGS can be overkill for the cost. Going forward, this might provide a targeted market for microarray usage in smaller experiments. It might also be the case that microarrays can find a leading role as a screening utility when a low-cost solution will suffice. Besides small genomes and low cost solutions, there is also the potential for microarrays to serve a more complementary role to next-generation sequencing. Roche's Seq-Cap technology uses Nimblegen arrays to hybridize and enrich sequence. This enriched sequence is then used on Roche's 454 sequencer to produce sequence reads. It is quite possible that other uses of microarrays as a precursor to NGS will evolve, creating a new place for microarray technology in research protocols.

On the other hand, as the price of NGS drops and the data analysis and management issues are addressed, high throughput sequencing technologies will become even more accessible to researchers. In light of the ever-forward marching progress of genomics research, it seems that only time will tell what new roles microarrays will fill in research.

• Michael F. Smith
• Co-founder

References:

1. Wikipedia - DNA Microarray
2. Nature Methods - The beginning of the end for microarrays?
3. Briefings in Functional Genomics - Advantages of next-generation sequencing versus the microarray in epigenetic research.
4. Illumina HiSeq 2000 Press Release

Thanks to all the attendees that welcomed us at our first Plant and Animal Genome Conference (PAG). We had a lot of fun meeting everyone and discussing some exciting research topics. We certainly enjoyed the San Diego weather as well! We look forward to attending PAG again in 2011.

Is next-generation sequencing making microarray technologies obsolete?

• Absolutely, we don't even consider using microarrays anymore!
• No, microarrays are still an important tool.
• Each has their own place and advantages in research.
• A symbiotic relationship will develop between the two

Recommended readings from the Bio::Neos team:

• Bioinformatics challenges for genome-wide association studies. by Moore J, et al. Bioinformatics. 2010 Feb 15;26(4):445-55. Epub 2010 Jan 6. (PMID: 20053841)
• Common variants near TERC are associated with mean telomere length. by Codd V, et al. Nat Genet. 2010 Feb 7. (PMID: 20139977)
• BFAST: an alignment tool for large scale genome resequencing. by Homer N, et al. PLoS One. 2009 Nov 11;4(11):e7767. (PMID: 19907642)