Skip to main content
Premium Trial:

Request an Annual Quote

Microarrays Make the Cover at AACR Meeting: Data Analysis, Dx Discussed


From a conference guide that features microarray spots on its cover, to Todd Golub’s speech at the opening plenary session, to dozens of educational sessions and poster displays, microarray technology has been a strong theme throughout the 95th annual meeting of the American Association for Cancer Research this week, at Orlando’s sprawling Orange County Convention Center.

A green-and-black circle, containing an array of familiar ‘homebrew’ microarrayed spots, ablaze in fluorescent Cy3 dye green [with a few malformed ‘donut’ spots], provides the background on the front cover of the meeting’s inch-and-a-half thick proceedings guide, ancillary program, and proceedings supplements handed out to each of the estimated 15,000 cancer researchers in attendance.

The technology is fairly straightforward, but analysis is the big issue, said Golub of the Dana Farber Cancer Institute and an associate professor of pediatrics at Harvard Medical School, in the opening plenary on Sunday morning.

In microarray data analysis, “nearly everyone is immediately brought to their knees in trying to understand what these gene-expression signatures mean,” he said.

“One often has one of two problems. There is either a long list of genes that is rather anonymous and it [is] difficult to understand what the biological story is trying to tell you. Or, there aren’t actually any genes that are statistically significantly differentially expressed, and that [result] doesn’t match with our current understanding of the problem.”

Of course, this approximately 10-year-old technology has a few more well-documented issues than just interpreting the torrent of data derived from the analysis of just a single microarray chip, but Golub said he is seeing progress. He cited the Gene Set Enrichment Analysis, which was recently produced by a postdoc in his lab, and involved mining meaning from microarray experiments where individual differentially expressed genes were measured at a level statisticians would call insignificant. When viewed en masse, through this pattern recognition process, the differences suddenly become “extraordinarily statistically significant,” he said.

Progress like this, he said, is “reassuring to all of us hoping that genomic information will be comparable and will tell us the same biological story.

“We are becoming increasingly convinced that this is indeed the case.”

And, in the case where “God forbid, one would want to bring clinical implementation, this pattern recognition is a good argument for not reducing these gene expression diagnostic signatures to single diagnostic genes, but using the genomic power of multiplexing these signatures and incorporating these studies with other studies we are comfortable with in the lab.”

Therein lies another problem: studying the protein products of these genes, where too few antibodies work well for immunohistochemical examination, he said.

Many cancer researchers, like Abebe Akalu, a post-doctoral research associate of the Medical School of New York University, who is focusing on extracellular matrix proteins, have dabbled in microarray analysis and are impatient to move to this next step.

Akalu said he has conducted microarray analysis with the Affymetrix platform for a year, using about 20 GeneChips at a cost of $500 each, to zero in on some 300 genes of interest.

“We have in hand some very interesting genes, and we want to look at the protein level using Western blots, ELISA, and in situ hybridization,” he said. Additionally, he said he plans to use another dozen GeneChips, performing experiments in triplicate to validate the data. He plans to use Affymetrix’s new single-array, human whole genome chips at a cost of $600 each.

The work has already yielded tangible results, as the school has licensed the markers and applied for a patent, Akalu said.

The technology has made its way into the labs of some of the foremost cancer researchers, including Judah Folkman, director of surgical research at Harvard Medical School Children’s Hospital, and a pioneer in the study of anti-angiogenesis treatments for cancer, and Paul Talalay, professor of pharmacology and molecular sciences at the John Hopkins School of Medicine, whose lab is perhaps most famous for its identification of sulforaphane, found in broccoli and broccoli sprouts, as an anticarcinogen.

Both researchers told BioArray News that their labs are not conducting microarray-based research directly, but are happily using data derived from array-based analysis.

But don’t rush to celebrate the coming of age of this technology quite yet. Daniel Hayes of the University of Michigan Comprehensive Cancer Center asked: “Do genomics and proteomics change the rules?”

No, he said in addressing an educational session on Saturday.

While molecular technology has allowed researchers to identify markers that could be important in the study of cancer, “Good technology is not good science,” he said. “Clinical science is just as important as the technological science. It should be hypothesis-driven, or hypothesis-generating. You have to have appropriate study design, appropriate controls, and careful analysis.”

Reliability and reproducibility —two major hurdles microarray technology must conquer— are still required to make good science.

Still, he said, in an event with perhaps as many as 100 talks featuring at least one microarray slide, he felt he had to display a gene-expression array slide too.

He flashed a slide of the tartan of the MacLeod clan of Scotland, with “bad” written underneath it, followed by the Hay clan, his forbears, with “good” written underneath it.

“I’m Scottish,” he said. “And it occurs to me that this is just a Scottish approach to science, where you can recognize the bad clans from the good by their tartans.”



The Scan

Researchers Compare WGS, Exome Sequencing-Based Mendelian Disease Diagnosis

Investigators find a diagnostic edge for whole-genome sequencing, while highlighting the cost advantages and improving diagnostic rate of exome sequencing in EJHG.

Researchers Retrace Key Mutations in Reassorted H1N1 Swine Flu Virus With Avian-Like Features

Mutations in the acidic polymerase-coding gene boost the pathogenicity and transmissibility of Eurasian avian-like H1N1 swine influenza viruses, a PNAS paper finds.

Genome Sequences Reveal Evolutionary History of South America's Canids

An analysis in PNAS of South American canid species' genomes offers a look at their evolutionary history, as well as their relationships and adaptations.

Lung Cancer Response to Checkpoint Inhibitors Reflected in Circulating Tumor DNA

In non-small cell lung cancer patients, researchers find in JCO Precision Oncology that survival benefits after immune checkpoint blockade coincide with a dip in ctDNA levels.