Bioinformatics lies at the heart of the broad personalized medicine collaboration that GE and Celera announced to much fanfare last week. The companies disclosed an initial project to develop new imaging agents that will selectively target certain cancer-associated proteins, but a second collaboration based around bioinformatics will play a critical role in that initiative — as well as several others that are still under discussion, according to informatics executives at both firms.
John Reynders, vice president of informatics at Celera Genomics, told BioInform that the bioinformatics project “took on a life of its own” during initial discussions between the two companies, and that the algorithms and tools that come out of the informatics collaboration should “touch upon a lot of other projects that are under consideration” between GE and Celera.
Brion Sarachan, bioinformatics lab manager at GE Global Research, agreed. “The intent is for the bioinformatics work to go hand in hand with the initial broader project in imaging agents, but with future projects as well,” he said.
When Celera and GE sat down to discuss their collaboration, it turned out that the global technology conglomerate and the biotech best known for sequencing the human genome had a great deal in common when it came to bioinformatics. Celera — which has moved beyond its genomics roots to embrace proteomics, expression analysis, clinical data analysis, and other experimental methods to support its therapeutic development goals — has been developing a suite of bioinformatics tools to integrate and analyze these large, disparate data sets. GE Global Research, meanwhile, launched a bioinformatics effort several years ago [BioInform 07-04-03] to pull together heterogeneous data sets to support its own research on molecular imaging-based diagnostics.
“Something that really resonated early on in the GE/Celera discussions was how both companies had this underlying heterogeneous multivariate data analysis challenge,” Reynders said. Even though each company was merging different kinds of data, “the underlying challenge was similar, so the thought was [that] this is an area where we can pool our ideas, pool our talent, and maybe even find ways of combining the kinds of data we’re looking at to come up with a way of understanding what the data is telling us better than either company could by itself.”
Celera brings to the collaboration a mix of genomic data, genotype data, polymorphism data, mRNA expression data, and clinical data from its internal drug discovery effort, while GE Global Research has a blend of data from its imaging instruments along with clinical information in a number of disease areas. “There might be companies that are doing this kind of multivariate analysis in an imaging space or doing the multivariate analysis in a proteomic or genomic space, but when you combine it, that is what I think is truly unique,” Reynders said.
Early, but Optimistic
Officials from both firms stressed that the collaboration is in its very earliest stages, so the ultimate payoff from blending this wealth of data remains on the distant horizon. Nevertheless, they know where they’d like to end up. The companies are setting out on an informatics development path that will enable the parallel discovery of diagnostics and therapeutics using combined data sets — an approach that Sarachan deemed “pretty unique.”
“We want to extend our analyses to mine the same data sets for potential diagnostic markers and potential imaging reagent targets,” said Vaibhav Narayan, associate director of informatics at Celera Genomics. “There are different ways of looking at the same data from those different perspectives, and we hope to tailor this informatics collaboration to make sure that we are able to find both diagnostic and therapeutic value from our data.”
Sarachan added that the partners consider “co-development between the therapeutics and the [molecular imaging] diagnostics” to be a key goal of the collaboration. GE is betting that molecular imaging will play a large part in personalized medicine in the future, but it needs the molecular biology know-how — and data — from a partner like Celera to drive that vision forward. “As the data is pooled and analyzed together … it may be that the biomarkers for a drug and the biomarkers for molecular imaging may be not the same ones, but very closely related ones — perhaps on the same pathway. So it may be a matter of looking at biology at the mechanistic level together and determining what is the basis for a drug, and what is the basis for a very closely related imaging test,” he said.
From GE’s perspective, “it’s always been on the horizon for us to think more about genetics, genomics, and proteomics,” Sarachan said. Now, he added, after GE’s acquisition of Amersham in April to create the GE Healthcare business unit, “that is absolutely within our interest, and we think there are unique things we can do with Celera in combining their unique genetic and genomic data with the clinical data we’ve been working with and some of the techniques for the life science data that Amersham has developed.”
Sarachan said that GE Global Research was in discussions with Celera regarding this collaboration well before the acquisition of Amersham came into play. While the scope of the collaboration thus far is between GE Global Research — the company’s R&D arm — and Celera, extending the effort to include Amersham’s informatics expertise on the business side of the company has not been ruled out. Sarachan said that his group at GE Global Research has been “building relationships and bridges and collaborations within what had been Amersham, and which is now part of GE Healthcare, and there may be other pockets of expertise and other people we could potentially bring into this, although at this point we haven’t broadened it to that yet.”