Predictive
Informatics, the hopeful title of a session at lastmonth’s Drug Discovery Technology & Development World Congress,remains an enticing but mostly elusive goal. The proliferation ofinformatics software, however, is entirely concrete, as speakers fromNational Cancer Institute (
NCI), Novartis, and Virginia BioinformaticsInstitute (VBI) demonstrated in talks about the
tools they’vedeveloped, followed by a panel on
systems biology’s prospects.Indeed,the range of tools developed at NCI and presented by John Weinstein,head, genomics & bioinformatics group, laboratory for molecularpharmacology, was astonishing, and they are all freely availablethrough NCI (discover.nci.nih.gov). No wonder the commercial informatics world has struggled.Evenwith its wealth of tools, Weinstein characterized the currentenvironment as “Wild West days” for technology to analyze large datasets. William Egan, a
computational chemist from Novartis, provided aglimpse into that company’s tools for predicting toxicology, and DariusDziuda reviewed a suite of tools being developed at VBI. Clearly, thesetools can help researchers make sense of unwieldy
data sets, thoughthey are often used singly and frequently defy easy integration withresults from other tools.In broad terms, systems biologyattempts to integrate various omic data — frequently incorporatinginput from informatics tools — to produce a holistic view of how agiven biological system works. But questioned by moderator Alan Louie,research director for Health Industry Insights, the systems biologypanelists demonstrated there is no consensus on how far systems biologyhas progressed, which part of drug development it’s most likely toimpact, or what the right approach is to creating these models ofliving systems.“As I think of system biology, there’s kind atension between this notion that (Leroy) Hood (founder of the Institutefor Systems Biology) and others sold us in the early 2000s, that youcould look at everything and put it all together,” said panelist StanLetovsky, senior director of computational sciences, MillenniumPharmaceuticals, “versus the picture that pharma has of traditionallyknowing a whole lot of your target and the molecule that’s interactingwith it and maybe a little bit about what it does to the system becauseyou’ve got to drive this thing down the pipeline as quick as you can.It’s not clear how that tension is being resolved.”“I don’tthink we’re seeing a lot of value in the short term for largequantitative models. They are just too unreliable at the moment. Sowe’re probably going to have more luck with the more correlative datakicking out the occasional target by hypothesis-driven or data-miningcandidate identification approaches,” he said.Bruce Gomes, headof modeling at Pfizer’s systems biology group, saw things a littledifferently. He allowed, “Systems biology is useless unless it actuallyanswers practical questions. I understand you said that large modelsare limited, and I think that is true. I think, however, the ability toanswer very specific drug discovery questions, where you can get smallamounts of information from very discrete experiments, will have animpact today. However, the ultimate goal is to give context to targets,to tell us all there is to know about it someday, and safety, what’sthe best delivery, what’s the dosing, where’s the best targeting,that’s going to happen some day.”
More abstracts about the Informatics Cornucopia part(1)