Applying Systems Biology to Benefit Human Health forum: topic
This is a public forum
General Feedback
Giulio Superti-Furga
Thursday, 09 October 2008 18:19 UTC
Frustrated from the skepticism one encounters within institutions, biopharmaceutical companies and funding agencies regarding the usefulness of systems approaches to medicine and in particular in drugs discovery and mindful of the danger of letting hyperbolic general promises create unbelievable expectations and delusions in this area with genomics-like anticlimactic conservatism as consequence, we have initiated a process that should lead to a community-wide research roadmap proposal. The purpose is to define the areas worth focusing on to obtain the facts that are needed as proof of concept for the entire community. In a careful process, we have engaged some of you first in interviews, then with on-line questionnaires and finally with a meeting restricted to 25 people that occurred in May (in Portofino, Italy). The outcome of the meeting is summarized in a Commentary that just appeared in Nature (enclosed). So far it has not been practical to engage all of you directly but we now need your help as we are preparing a WHITE PAPER that covers much more ground of the Portofino discussions and following that measures for implementation.
Here we look for general feed-back. We got this rolling and would like to know if the community is willing to participate in the autodiscipline needed to achieve the goals of the recommendations. Moreover, we are looking for further focus and detail on the proposed ideas and on what could be additional ideas that the group may have overseen.
Moreover
We will post the following topics:
Setting Data Standards
Optimising the application of existing tools
Predictive Toxicology
Therapeutic Area Focus
Communication and Outreach
Updated 09 October 2008 18:23 UTC
-
Replies
Jump to resultsResults
-
Setting data standards in the field of predictive toxicology has been a focus of the DrugMatrix program originally developed by Iconix now taken over by Entelos (www.entelos.com) and a description can be found at http://www.iconixbiosciences.com/products/Entelos%20Toxicogenomics%20Products%20Overview.pdf
Although they were not calling this approach “systems biology”, clearly this was a systems biology approach to tackle (amongst other issues) predictive toxicology.
Thus I suggest this new initiative should know about what has been achieved in this program already.
-
Uli, thanks. You are perfectly right. Saroja Ramanujan of Entelos was one of the “Portofino 25” and Alex Bangs was involved early in the process. Waht we mean here is that we need a coordinated effort, convincing the biopharma industry to contribute at least part of their (historical data and new standardized.
-
FYI, the link to the article being discussed is here. It is a Commentary published in the Oct 9 issue of Nature.
-
I completely agree with the analysis of Henney and Superti-Furga on the need to intensify the involvement of systems biology approaches in drug discovery. Recent advances in the assembly of drug-target networks, disease networks as well as in the methods to predict missing network links and nodes see our recent summary in the Journal of Biology make this approach even more promising.
-
I liked the subject of this commentary. I think part of the demystification of Systems Biology is related to coming up with realistic things that can be done now that have an impact on human health. I think only then can one begin to justify the more blue-sky type approaches (whole organism modelling, etc.).
I think the general power of networks needs to be more explicitly highlighted. Specifically, that networks can approximate the biological system to some extent, even when most of the nitty-gritty details (biophysical constants, etc.) that exact models need are missing. The pioneering work of Mark Gerstein (Science 2002) showed that networks can be used to predict missing edges (i.e. interactions), and this very concept has since been used for a number of things, including our own attempts to predict phosphorylation sites successfully (Linding et al, Cell, 2007).
On predictive toxicology, there are already initiatives underway. I’m involved, for instance, with Cambridge Cell Newtorks, which has developed software to predict toxicity based on biological networks. I would encourage anybody interested in this area to take a look at them.
Rob Russell
Group Leader, Structural Bioinformatics
EMBL, Heidelberg -
Thanks Superti-Furga and Henny – your effort will go a long way in furthering the cause of systems biology. Indeed, the ultimate aim of systems biology is to delineate and comprehend the functioning of complex biological systems in such details that predictive models of human diseases could be developed. Due to immense complexity of higher organisms, systems biology approaches are however currently focused on simpler organisms. Amenable to modeling, these simpler organisms may offer a unique opportunity to dissect the pathophysiology at systems level, through reiterative processes of hypothesis generation and hypothesis testing.
I would like to add that neuropsychiatric disorders and neuroactive drugs should also be brought to focus. For example, consider epileptogenesis that is relevant in different neuropsychiatric conditions in addition to epilepsy (similarly, relevant in development of antiepileptic drugs which can also be used in treating different neuropsychiatric conditions). As inherent complexity of mammalian brain does not render the available rodent epilepsy models as amenable to systems modeling, a simpler organism model may be of immense potential here. In this direction, we have recently developed (unpublished) a Drosophila predictive systems model. Three poster abstracts (nos. 553, 575 and 577) in the recently held Human Genome Meeting at Hyderabad (HGM2008; http://hgm2008.hugo-international.org/Abstracts/Publish/WorkshopPosters/) provide a glimpse of our model. In brief, our Drosophila systems model is expected to be valuable in identifying disease, drug target, biomarker and pharmacogenomic candidates, and in screening of potential therapeutic agents.
Cumulatively, we strongly feel that systems biology has immense potential in human health. Pharma industries, funding agencies, publishers and peers all need to understand this and act accordingly.
-
Dear Drs. Henney and Superti-Furga:
Francisco Azuaje made a thoughtful comment, which I agree. Here, based on my nearly 10 years experience in systems biology, ranging from wet to dry labs, from applications to basic biology, I would like to share a few thoughts, and adding some points.
Yes, I believe the network thinking is THE way to go. Here are two our approaches at the endogenous and metabolic networks, two hierarchical layers minimum for drug discovery and development purposes.
a) Towards Kinetic Modeling of Global Metabolic Networks: Methylobacterium extorquens AM1 Growth as Validation,
P Ao, LW Lee, ME. Lidstrom, Lan Yin, and XM Zhu,
Chinese Journal of Biotechnology 24 (2008) 980 – 994.
http://arxiv.org/ftp/arxiv/papers/0808/0808.0220.pdf
b) Cancer as Robust Intrinsic State of Endogenous Molecular-Cellular Network Shaped by Evolution. P. Ao, D. Galas, L. Hood, X.-M. Zhu,
Medical Hypotheses 70 (2008) 678–684.
http://dx.doi.org/10.1016/j.mehy.2007.03.043My experience shows to me that a few agents, or too few a pathways, i are nearly impossible to understand complex diseases such as cancer. DNA sequence or gene analysis alone will not lead us any further, either. Fortunately, this latter point of view begins to become a mainstream, showing by a recent group of high profile papers in Nature and Science.
One main feature missing in the discussion so far is a general theoretical framework.
My impression is that, despite the name “systems”, most approaches proposed so far in systems biology are of receipt type. While they are important, we need an understanding on the mechanisms of complex biological phenomena on a global level. Such understanding can of course greatly enhance our ability to discover and to develop new drugs. I have raised this theoretical framework issue recently in a broader perspective:
c) Borges Dilemma, Fundamental Laws, and Systems Biology,
P. Ao, Bioinformatics and Biology Insights 2 (2008) 201-202
http://la-press.com/article.php?article_id=693In the following, let me make some comments on the suggested topics.
Setting Data Standards—This is absolute important if we wish to understand each other and to avoid unnecessary mistakes, when the network becomes large. Our criteria for standardization are: sound biological, chemical, and physical base, and, easy to use.
We have some initial success here on metabolic networks:
d) Generic Enzymatic Rate Equation under Living Conditions,
Lee, Yin, Zhu, and Ao, J. Biol. Syst. 15 (2007) 495-514. http://ejournals.wspc.com.sg/journals/jbs/15/1504/S0218339007002295.htmlOptimising the application of existing tools—Indeed this is important. Above d) can be regarded as an example of optimization of existing tools. But, it is not enough. New tools need to be developed.
We have been doing this, too, as indicated in a) and b).Predictive Toxicology—Good framework is predictive. Our above on kinetic modeling of metabolic network, a), has this feature.
Therapeutic Area Focus—The endogenous network modeling on cancer genesis and progression has a few very interesting predictions connected to this, even at our initial modeling stage.
This can be regarded as another important evidence for network thinking and modeling.Communication and Outreach—There is a need to design suitable courses to educate next generation biologists. In my view, evolution theory should be the unifying foundation. Unfortunately, there have been a great deal of confusions among biologists. For example, what the natural selection really means is still a subject of controversy.
Many biologists only pay a lip service to evolution.I will be happy for further discussions ( aoping@u.washngton.edu ). P. Ao .
-
One of the points I would like to highlight in Giulio’s response to my comment (posted to topic section “On the paper A network solution") is that “we need ideas on how to… impress the biopharma community really”.
Within a long-term vision or framework, we could also consider ways to “impress” by accomplishing possible short-term or near-future goals.
Would it be possible to do this by demonstrating that in 5 years, for example, significant translational research can be done?
My next question is: How significant? Would it be possible to impress as early as possible without directly offering a new treatment?
Could we impress by offering new diagnostic kits, including independently validated diagnostic signatures? Or perhaps by demonstrating the value of similar outcomes in a clinical trial context, e.g. potential applications as surrogate endpoints? Or by offering a specific set of “druggable” targets?
It is evident that we want to go beyond that. I just wonder if these ideas can find a place within the “near-future” agenda.
-
I am happy to see that standardisation of functional enzyme data was addressed at the Portofino conference. I also noticed with great interest that there is a demand for the establishment of a consortium which should be concerned with standardisation. Moreover, I also read that there seems to be the wish for guidelines for reporting enzyme data.
In conclusion, some of these demands are already going to be fullfilled. This is to inform you that the STRENDA commission website has set up guidelines for standardised and comprehensive reporting of enzyme data. These guidelines are more and more adopted by biochemistry journals e.g. _JBC _and Biochemistry.
The commissions’s work is accompanied by a conference series called ESCEC (Experimental Standard Conditions of Enzyme Characterisations). This conference is the platform to discuss all aspects and requirements of standardized data.
Please let me know if you are interested in further details of this issue.
-
Someone mentioned the need to impress the pharmaceutical companies. It will be difficult to find a clear cut example of a systems biology drug on the market, and it will take time to see that anyway. The industry is looking for new ideas and willing to explore the possibilities given by a systems biology approach. It will be a trip, and we will discover the value of systems biology by applying it. It is however important to avoid unrealistic promises and keep a sober approach to avoid to generate frustration (as Giulio pointed out).
The article by Giulio and Adriano was very nicely written. The link between personalized healthcare and systems biology has not been elaborated as much as I had wished due to editorial limitation. The focus of systems biology on multiple mechanisms (think of safety and efficacy in general at the easiest level), the development of unprecedented technologies for discovering and monitoring biomarkers and the promise to link biomarkers to the mechanism of action of drugs, all point in the direction of personalized healthcare solutions. In the future we could talk of “predictive” treatments. In my opinion this view can be already now concretely implemented by the industry using a systems biology approach (there are few examples). It does not need to be perfect, but it can be better than the current “one fits all” drug discovery model. Changes in paradigm occur over long time frames, but already now systems biology can respond better than what we currently have to one of the most challenging needs of the industry, personalized medicine.
The value of systems biology should be communicated to industry (and payers) using personalized healthcare solution examples (including biomarkers and combination therapies).
I would be glad to discuss this more in detail.
Results
-
