Skip to main content
Premium Trial:

Request an Annual Quote

Roundtable Systems Biology, Pharma Style


By Meredith W. Salisbury


In our final systems biology roundtable, three experts with pharmaceutical backgrounds talk about the challenges of getting these heavyweights into the systems bio arena. Are pharmas up to the task, or are skeptics right that these mega-organizations just aren’t nimble enough for an interdisciplinary, non-silo approach to drug discovery and development?

Our experts brought different perspectives. Jack Beusmans, an informatics scientist in the pathways group at AstraZeneca R&D, has a background in biology, mathematics, and computer science. Norman Lee, whose PhD is in pharmacology, has spent his career in the public sector — the last 10 years of that at TIGR, where he studies functional genomics. And Jim Xu, an associate research fellow hailing from Pfizer Global Research & Development, has been working in toxicology for the last decade.

The following pages include excerpts of their discussion, which took place in Boston this August.


Genome Technology: Pharmas are certainly pouring money into this concept of systems biology, even though at this point the concept itself is still pretty amorphous. What is the perceived benefit of systems biology, and what makes it so great that pharmas are willing to invest in it?


Jack Beusmans: It’s nothing earth-shatteringly new. Very specifically, people would like to know for instance what biomarker would be significant in clinical trials, and this is another way of getting at that. The hope is that we will get there quicker [and] with more valid biomarkers than we have had in the past. Hopefully we will answer all questions better and faster, and that’s the mantra.


Jim Xu: My understanding of systems biology is that a lot of the low-hanging fruit has already been picked, like single-target diseases. Now we are targeting more complicated diseases — within cancer, for example, there are different etiologies, different targets — where you really want to understand the complex interplays of these targets in the sense of pathway analysis.

This is a totally different ballgame from just culturing some bacteria or a fungus and trying to kill it. The systems biology perceived benefit is to understand the complicated pathways and which particular target or targets you should inhibit to be able to get clinical benefit. We also want to use systems biology to understand drug toxicity.


Norman Lee: That’s one of the big buzzwords now: toxicogenomics. We have these therapeutic drugs — some of them we have a good understanding of the mechanism of action, others we don’t have a good understanding, and there are also a lot of unwanted side effects. So from our perspective when we try to understand systems biology, we like to try to understand the gene networks potentially that are involved. How do we bring that information in terms of therapeutic effect, and so the hope is eventually that we can understand the signaling networks, the gene transcriptional networks, the protein networks. Can we identify critical branch points and then attack those from a therapeutic standpoint and try to get away from those unwanted side effects or those toxic effects?


GT: What are the key challenges in getting a systems biology effort underway?


Beusmans: One of the things that we have learned in our efforts over the years is that the quality of data is still very poor. Everybody likes to talk about the fact that we live in a data-rich environment, but one of the things that we have learned in trying to put it together in a coherent framework [is] that often that data is incomplete. In our case where we’re trying to develop mathematical models for a particular disease, we were forced to gather data from wherever we could find it. We realized that really the big bottleneck is good quality data.


Lee: Then really getting the computational people in house to be able to pore through that data, to try to distinguish the good data from the bad.


Beusmans: Where we are putting a lot of effort right now is to try to get a productive synergy between the modeling effort and the data gathering effort, where the model would actually be used to inform the next set of experiments. Our effort now is trying to use the model and engineering approaches about system identification to identify what is the most informative experiment that one could do next.


GT: What’s wrong with the data we have, and how do we get better data?


Beusmans: Well, there’s missing data. There’s very limited data on protein interaction networks. It’s only a limited view of what’s happening. We are currently just scratching the surface of, for example, tyrosine phosphorylation status — people would do things very ad hoc. We really have to go beyond that if we want to have a systems perspective where we can say we know what the phosphotyrosine signal is in this network in response to some external signal.

One big problem is the model systems that we have had to rely on: cell lines instead of primary cells or primary cells instead of tissues. That needs to be improved; things are moving in the right direction. Or develop better model systems where the cell is sitting in the relevant context rather than agar.


Xu: We have a lot of genomic data, but a lot of what Jack talked about is actually protein modification or protein interaction data. These kind of ¯omic technologies are starting to mature [but] of course there are a lot of data gaps in that area. And those are the most fundamentally important for medicine: a lot of small molecules, traditionally speaking, are either inhibitors or agonists of protein functions.


Lee: I think the other perspective is an understanding of the platforms that are being used for measuring. There is definitely good and bad data, [but] there are other times when there are just platform differences. Having an appreciation of the different technologies that are being used in these measurements is an important aspect to consider too.


Beusmans: In that sense it’s important to have computational biologists working very closely with the biologists. We cannot take data at face value. It’s hard because these are different cultures and it takes work, but it’s very important.


GT: What are the technologies that are most important to start with?


Lee: The most obvious for us is RNAi. We’ve sequenced all these genomes — it doesn’t matter what organism — there are so many holes in our knowledge base. Half the genes we think we know what they’re doing, and that’s just we think we know, and the other half we just don’t have a clue. I think with the introduction of RNAi we finally have a mechanism at hand to try to decipher what some of the functions of these genes are.


Xu: Systems biology will take technologies from various disciplines — one thing that I would like to mention is cell-based imaging technology, where you could observe what’s going on inside cells using either fluorescent proteins or antibodies to track a specific kinase. Location, location, location is very important.


Beusmans: That’s a very important area: high-content biology. AstraZeneca has quite a big focus on that. If you marry [RNAi] with a technology like imaging with specific antibodies, looking at phosphorylation signatures over time, I think that combination is very powerful. That would be my platform of choice if I were to set something like that up. Imaging techniques have shown that you really need to look at subcellular localization and how things change over time — that’s certainly the next step in this area.


GT: Let’s talk about the cultural challenge. Skeptics say that pharmaceutical companies won’t be able to get past the silo mentality to make systems biology happen. What do you think?


Xu: The pharmas have recognized that for quite a long time, so the old structure of a biology department as opposed to a drug safety department, for instance, may not be conducive to this kind of integrative science. They’re trying to put it all under the umbrella of systems biology, bringing in people who have significant contact with those traditional lines. That said, you still need the traditional lines for the single reason of conducting GLP studies to satisfy the government regulators. But in early discovery you can mingle as much as possible.


Beusmans: We can see that throughout the drug discovery process, which used to be a very linear process. It’s now becoming much more parallelized and there is more interaction between the various steps in the process. People are aware that the kinds of problems we are trying to tackle are very difficult and that we need to take drastic actions in some cases. Just chugging along as we have been doing is not going to work. I can see already within our company concrete evidence that things are loosening up and that people are willing to cross these traditional boundaries.


Lee: It’s still very important that we have these groups that are focusing in on a particular protein, if you will, but I think that’s changing with genomics in all departments — in pharmacology, in biochemistry, people are realizing they’re going to have to take a broader perspective. There are certainly advantages of seeing the whole genome, the whole function. In that sense even in academia there’s a shift in that direction.


GT: Are regulations from the FDA keeping up with the times, or are they holding you back from this kind of shift?


Beusmans: The FDA has issued some white papers lately where they do provide guidance and recommend more integrated biology and perhaps even use computational simulations. So in some sense the FDA’s actually been pushing industry ahead in that direction.


GT: We’re seeing a lot of the usual partnerships develop between pharmas and small biotechs in response to the need for systems biology, but also some unique ones between pharmas and academic labs. Is this a new trend?


Beusmans: That’s the approach we have taken as a group. We identified the Lauffenburger lab at MIT as the gold standard, essentially. They had an expertise in an area of biology that we were very interested in — the EGFR signaling pathway — so we teamed up with Lauffenburger almost two years ago. That has been our way of getting up to speed.


GT: How do you evaluate a potential collaborator?


Beusmans: It is ideal if you and your collaborator are in close proximity. When we were thinking about systems biology internally, we really felt that it was important to have both the computational and biology groups work closely together — and that’s exactly what Lauffenburger has been doing. Then [the expertise] has to be in an area where you have a need, which in our case was EGFR signaling. Lauffenburger has a lot of experience working with pharma and biotech, so he knows what we are looking for; that’s really important as well. Of course you look for quality first.


Xu: There are three very important points: depth of knowledge, practical thinking, and close proximity. You can’t stress that enough for the collaboration to stay on course.


Lee: Common interest, common goals, different partners bringing in different expertise. Our collaborators are spread out all over the place, but we have common interests so that always drives the research forward.


GT: What is it going to take for people to look at a pharma five or 10 years from now and say, ‘Wow, they’ve got a great systems biology team’?


Beusmans: The drugs. That’s the business we are in. We can have the best systems biology group in the world, but if it doesn’t lead to better medicine, then from a pharma perspective, that was a waste of money. What we need now are clear examples in which the discovery or development of a drug was guided by insights from a systems biology approach.


Lee: That’s an interesting perspective from pharma: is there a successful drug in the market? From the academia standpoint, the success comes from a better understanding of the biology.


Xu: As a pharmaceutical scientist we have to be very mindful of how our understanding of systems biology is being applied to either better targets or a better therapeutic index such that we can show some clinical benefit. That’s a matter of survival for the pharmaceutical company.



The Scan

Genetic Risk Factors for Hypertension Can Help Identify Those at Risk for Cardiovascular Disease

Genetically predicted high blood pressure risk is also associated with increased cardiovascular disease risk, a new JAMA Cardiology study says.

Circulating Tumor DNA Linked to Post-Treatment Relapse in Breast Cancer

Post-treatment detection of circulating tumor DNA may identify breast cancer patients who are more likely to relapse, a new JCO Precision Oncology study finds.

Genetics Influence Level of Depression Tied to Trauma Exposure, Study Finds

Researchers examine the interplay of trauma, genetics, and major depressive disorder in JAMA Psychiatry.

UCLA Team Reports Cost-Effective Liquid Biopsy Approach for Cancer Detection

The researchers report in Nature Communications that their liquid biopsy approach has high specificity in detecting all- and early-stage cancers.