Affymetrix vs. RNA-seq

Thursday, August 18th, 2011 by hinrich

Affymetrix GeneChips

As next-gen sequencing technologies mature and first approaches for third-gen sequencing (direct sequencing of the nucleic acid without the necessity to prepare a sequencing library) become commercially available, scientists and financial analysts wonder how Affymetrix will respond.

Beginning of this year Xu et al. published a paper with the title "Human transcriptome array for high-throughput clinical studies" in PNAS where they demonstrated the use of a new Affymetrix microarray design.

The main features of the array are the following:
• improved sensitivity to detect low-abundance transcripts
• cartridge format - not GeneTitan peg arrays
• not just gene expression but also profiling of exon-level expression, alternative splicing, SNP coding, drug-metabolism enzymes and transporters, and non-coding RNAs
• highly reproducible measurements
• no need for mRNA purification

In the article the team also draws some conclusions when they compare the array with RNA-seq. Well known shortcomings of current RNA-seq approaches are the low throughput, the high costs and the high demand of RNA - something that can be critical for studies with little sample material. While both platforms are shown to detect similar expression changes at the gene level, the array appears to be more sensitive at the exon level.

The author's recommended application of the array is to combine it with RNA-seq. In a first stage RNA-seq would be used to characterize the most relevant transcriptomic properties of a given disease. During the second stage thousands of patient samples are screened for those characteristics. This approach fits nicely with what I had discussed in my blog post from November 2008.

The article by Xu et al. can be found here.

Posted in Molecular Profiling