By John Russell
Sept. 25, 2008 | The attention being given to two recent Science papers* suggesting pathways, not genes, are better targets for solid cancer tumor treatment is good news for all pathway tool providers. Such tools were important in this work by prominent cancer researchers, which is discussed in the lead article in this issue of Predictive Biomedicine. The attention is especially sweet for GeneGo which provided tools and data analysis, and had two authors, Tatiana Nikolskaya (president and CSO) Yuri Nikolsky (CEO), on the papers.
It’s probably too much to declare these are heady days for pathway tool providers, but there is no denying their increasing traction inside pharma, their growing impact on research, and the soaring aspirations of the pathway tool provider camp. Currently, GeneGo, Ingenuity Systems, and Ariadne dominate the market, although there are many others. While the Science papers shone a flattering spotlight on GeneGo last month, all of the pathway tool providers are trumpeting the growing number of their citations in journals. Ingenuity, for example, touts 1000 publications in 300 journals on its website.
Whether by serendipity or shrewdness, GeneGo issued a new release of its MetaCore product suite (v.5.0, up from v.4.7) on August 22, which VP of business development Julie Bryant calls a major update. MetaCore has long been regarded as a powerful platform. Many of its new enhancements relate to ease of use and are likely to please both novice and experienced users (more details below). Virtually all pathway providers are racing to keep pace with the data deluge, improve interfaces, and build specialized portions of their databases around specific diseases, drugs, and toxicity.
Early last year, for example, GeneGo collaborated with the Cystic Fibrosis Foundation Therapeutics (CFFT) and launched MetaMiner CF, a disease-specific tool embedded in MetaCore. Ingenuity is working with the CHDI Foundation to create a similar offering around Huntington’s disease. The new GeneGo tool contains rich information about cystic fibrosis and represented an important expansion by GeneGo into disease-specific areas. It has since outlined a roadmap for more disease specific tools and has added staff for the development.
“I was in Moscow this year and we have within the company a big room of biologists, a big room of chemists, and big room of programmers, and also MDs on staff. They are all very disease-focused on oncology, metabolic disease, CNS [central nervous system], cardiac, immunology and inflammation, and we’re starting a new stem cell area,” says Bryant. “Oncology is going to come first, closely followed by CNS and immunology and inflammation.” GeneGo works with partners on these projects, but Bryant declines to say who just yet.
As you may infer, GeneGo’s roots are in Moscow where Nikolskaya and Nikolsky trained, and it has a big presence there. It was founded and headquartered in St. Joseph, Michigan, but now has several offices in the U.S. and abroad. The staff numbers over 120 and is growing quickly.
“We spent a lot of time building these multi-step canonical pathway maps, and we’ve got over 700 now. This is the largest collection in the world. Even if you put all the commercial and public domain pathways maps together, we’ve still got 60 to 70 percent more than anybody else. We’re going to build on that with these normal versus disease pathways and also mechanism and process pathways,” says Bryant, a vigorously competitive market watcher.
In characterizing the market, Bryant says adoption inside pharma and academia has been strong, though less so at biotech. She says demand for chemistry knowledge has been growing, which is something GeneGo has been delivering for some time. Making the tools easy to use, she says, remains the key to expanding their use, and she believes that is happening now. GeneGo’s near-term goal is building these disease-specific platforms and making sure it has good coverage of toxicology.
“The end goal is towards personalized medicine where pathway tools are used by physicians, making decisions on what drugs to give patients,” she says. GeneGo currently has a collaboration with Craig Webb ,director of the program of translational medicine, Van Andel Institute. That work is with cancer patients on determining “what potential cocktails could be given. It’s an iterative process with results (patient response) informing further analysis. “We have some good success stories,” she says.
It seems like a lot of companies have their eyes on the personalized medicine prize in its myriad manifestations. Biosimulation specialist Optimata, for example, has worked in the clinic helping to choose optimum therapies for cancer. So has Gene Network Sciences. Entelos has launched a website, MyDigitalHealth.com, to help physicians, payers, and possibly patients make decisions. And of course, the consumer genomics companies (23andMe et al.) have a personalized medicine play. While specific offerings differ, it’s not so far-fetched to think a couple of the companies could collaborate or combine.
Other services continue to be part GeneGo’s business model. “They tend to be [around] drug discovery and development and they are very diverse, from prioritizing drugs to actually creating orthologs that aren’t really out there in the public domain. We’re also building ontologies for companies,” Bryant says. There is internal R&D around diagnostics, but she declines to elaborate at present.
The recent MetaCore release may gain added notice because of the Science publications. During a demo, Easy Search – part of new section in MetaCore called Easy Pages – lived up to its name, making it a simple matter for example, to probe the biology for STAT1 and to conduct a variety of analysis.
“We’re now really going from casual user to advanced user, and so anybody, including my mum, could put in a gene, a protein, a compound, a drug, find information, and they can just stop there and use it for data mining or go deeper,” Bryant says.
A good deal of effort was spent fashioning “pre-built” modules to speed user activity and shorten time to results. New automated workflows, many developed with pharma partners, were also added, including some with advanced modeling capability.
In broad terms, GeneGo distinguished between maps and networks. Maps represent canonical multi-step pathways that have been confirmed multiple times in the literature. Networks, on the other hand, are typically developed from “small experiment data” and represent possibilities.
MetaCore contains a number of network generating algorithms that permit users to explore data. GeneGo has also added a variety of pre-built networks that represent specific processes or drug target networks or toxicity networks and disease networks, “where basically not just the canonical pathways, but also all other valid binary interactions are represented. What you see is a schema of a disease or a schema of all of the interactions of the constituents of that process,” explains Ally Perlina of GeneGo, during the demo.
Visuals have also been enhanced. For example, working with a pre-built network of inflammation and interleukin signaling to a STAT, just mousing over a network element causes all upstream connections to light up yellow and downstream connections to light up blue. It’s easy to continue downstream to one of the lit up genes and then keep navigating through this network to see what the signaling cascade may become.
The new modeling workflows, “allow you to basically predict or model all of the pathways that lead, for instance, from transcription factors to receptors in your dataset.” In one example using an estrogen tamoxifen study, the modeling workflow found “all of the transcription factors in a database that lead to receptors in a user’s gene list. So, that’s something that is very useful because sometimes transcription factors are really transient and they’re not readily picked up on the arrays, so they may not have significant readings themselves or significant P-values,” says Perlina.
*Williams Parsons, D. et al. “An Integrated Genomic Analysis of Human Glioblastoma Multiforme.” Science, published online 4 September 2008
Jones, S. et al. “Core Signaling Pathways in Human Pancreatic Cancers Revealed by Global Genomic Analyses.” Science, published online 4 September 2008
This article first appeared in Bio-IT World’s Predictive Biomedicine newsletter. Click here for a free subscription.