Plectix Blog

Moving to the Cloud

By Vy Hoang

With the recent interest in cloud computing, it isn’t surprising that big pharma companies, such as Pfizer, Eli Lilly and Johnson & Johnson, are also experimenting with the cloud. Cloud computing offers many advantages for a low price. These companies can easily predict the time spent and the cost of a job on the cloud if they know what job is pushed into the cloud. Thus, the pharmas can manage their money and time in a more cost-efficient way. In addition, cloud computing is fast because of the access to a vast number of parallel CPUs. Pfizer reported that running their Rosetta macromolecular modeling took only 3.5 hours using 500 of Amazon’s Elastic Compute Cloud (EC2) instances instead of 48 hours on their internal computers. The New Jersey based Johnson & Johnson is hosting analytical software such as NONMEM (non-linear mixed effects modeling) in the cloud. Taking advantage of the on-demand cloud infrastructure, Johnson & Johnson is able to scale up or down to run large simulation jobs at will. Eli Lilly is running both non-research and discovery applications in the cloud, a few of which could not run internally. Therefore, the cloud can provide more efficient, scalable servers to run jobs that couldn’t be run in house.

Although the results look promising, cloud computing poses some challenges for pharma companies. The main issues are security and accessibility. Some people view the cloud as an unsecure space and mistakenly feel that cloud resources are just extra servers that Amazon is not using. However, as they learn more about how the model works with its on-disk encryption, in-memory encryption and wire encryption, the people are beginning to come around. Yet, the complexity of cloud computing is an additional factor that is making cloud computing difficult to use. Without the help of an IT team, the pharmas would have considerably difficulty in understanding how to use the cloud. Pfizer stated that it was hard to get metrics out of the cloud such as what jobs are running or how long they are taking without any IT help. Whatever’s going on in the cluster is not as simple as you think. The pharmas simply want the cloud to be easier to use with less technical complexity so the users can do what they need to do. Even with these challenges, pharma companies are not deterred because the vendors are giving in to their demands, making it more manageable to use the cloud.

These early cloud adopters believe that cloud computing is the next step in improving the quality of research. Researchers will be able to solve problems in the cloud that were considered impossible before. As long as you know what you’re getting into, cloud computing can be the right way to go.

Biological Logic

By Traci Freed

Cellucidate was recently in a Nature feature about intuitive approaches to computer modeling for biologists.  The article gives an interesting discussion of "executable biology", a new approach to biological modeling that makes simulations of cells easier for scientists to build, understand and study.  To read the full article, visit http://www.cellucidate.com/26nov09.html.

Getting Over the Hump

By Isha Antani

For many experimental biologists, the field of computational biology is a vast and unexplored terrain. Sure, there is curiosity about this exotic land, especially now with the increasing overlap between computational biology and the rediscovered systems biology. But, when asked to take the opportunity to embark on an expedition to explore computational biology, most are not interested.

It is not that they do not appreciate what the field has to offer. It is more that the community has had success through its traditional laboratory techniques and corresponding tried-and-tested analysis tools. Trained from their university days to utilize tools such as Microsoft Excel to digest data sets generated from Western blots and such, most experimental scientists cannot find strong reasons to get into a whole new field.

Furthermore, there is a distinct disconnect brought on by computational tools. Computational biology, which is a convergence of three distinct concentrations — biology, computer science and mathematics — often utilizes tools that require programming experience and is thus favored by biological scientists with solid foundations in some aspect of programming. Experimental biologists not trained to approach science from a programming perspective often find computational tools tasking, requiring precious time to learn and master. Additionally, translating laboratory data into the unfamiliar and unfriendly languages used by computational tools (e.g., ODEs, scripts, etc) is a tricky hill to surmount.

For computational biology to be properly integrated into the mainstream scientific process, there need to be better ways to merge the existing synergy between current biology and the new computational arm. Learning new techniques and transforming schools of thought requires time, which many scientists do not have. Perhaps then, the best way to better integrate computational biology with traditional laboratory approaches is to incorporate some computational curricula at an earlier stage in a scientist’s education. This can be done, for example, during the days of undergraduate studies, when no particular methodology has been engraved into the routine and the fertile mind is in the stage of gaining an understanding of the many scientific fields.

Systems Biology in a Nutshell

By Joshua Havumaki

As proof of concept applications such as Merrimack’s MM-121 (an ErbB3 inhibitor) slowly make their way into the mainstream, the systems biology community is continuing to grow and people are starting to appreciate the necessity of computational methods to model and explain complex biological systems. 

Within the community, computer scientists and mathematicians are rapidly developing methods for almost every step of the modeling process. Experimentalists are starting to dabble in small representative models consisting mostly of a few reactions or differential equations. In addition, certain community standards like SBGN (Systems Biology Graphical Notation) and SBML (Systems Biology Markup Language) are evolving and improving. However, in my mind, there are still a few problems that elude the explosion of the community and prevent it from unifying.

Computer scientists and mathematicians are developing methods to improve and/or redefine the modeling workflow. A quick browse through the CMSB (Computational Methods in Systems Biology) 2009 program provides a cross-section of methods ranging from the very theoretical (e.g. Computing Reachable States for Nonlinear Biological Models) to the directly applied (e.g. Improved Parameter Estimation for Completely Observed Ordinary Differential Equations with Application to Biological Systems). One major issue seems to be that most methods presented use a limited pool of applications and examples (e.g. the Raf/MEK/ERK cascade) and therefore rarely come to life for the biological community.

Additionally, it is a great undertaking for the average experimentalist to build an accurate, comprehensive model of a biological system. Commonly, experiments are not designed with specific mechanistic questions in mind. An experimentalist may ask whether a protein “activates” another protein, while a modeler must know if the interaction is direct and how the activation occurs. There is also a lack of specific parameter and protein abundance information in the literature, which is necessary for building biologically accurate models. Since experimentalists mostly think about biological systems from a higher level of granularity (e.g. not detailed mechanistically) the learning curve to adopt these methods for them may be too great to bear.

Standards like SBML and SBGN have yet to achieve widespread adaption by the community. One reason is the actual necessity and design of these standards don’t address the major pitfalls encountered by those trying to model complex systems. For example, SBGN doesn’t scale well with size and complexity.  This often results in trading comprehensively and accuracy for usability and understanding.

Lastly, biology, mathematics and computer science all have biases concerning scientific method and experiments. In math and computer science, even the most theoretical and abstract concepts can be explored while theoretical biological questions are impeded by money, time and technology. These problems, in conjunction with a lack of consensus on the definition of systems biology, result in a field that is fragmented and is yet to come out of its shell.

From Databases to Models?

By Ty Thomson

There are currently hundreds of online resources for biology, including countless databases containing information on proteins, protein-protein interactions, signaling pathways, and other biological processes.  In a recent perspective in Molecular Systems Biology, Bauer-Mehren et al. asked how easily the information from these databases can be harnessed and combined to produce a more complete picture of the cell, and perhaps even to produce computational models.  They found that although the information in many pathway databases is accurate and relatively comprehensive, there are still many challenges to merging information from multiple resources to construct computational models.

One challenge that they identified is that the scope of models greatly depends on the purpose of the model, therefore the modeler must carefully decide on the model boundaries in order to know what information is pertinent.  However, in order to decide on the relevant model boundaries, the modeler must know the relevant information.  This catch-22 means that information retrieval must be done in concert with evaluating the information in order to decide on appropriate model boundaries and to construct a useful model.  For example, while Bauer-Mehren et al. were able to automatically extract a number of signaling components and reactions in the EGFR pathway from the Reactome database, many components and reactions that the authors expected were not retrieved.  They had to make additional database queries to complete the model that they were constructing.  In this case, Bauer-Mehren et al. had carefully defined the boundaries of the model ahead of time by attempting to recapitulate the contents of a previous model.  If they had been constructing a model from scratch, they would not necessarily have known what components or reactions were missing without attempting fishing expeditions of additional queries.

Another challenge faced by the authors is that information in these online resources is often at differing levels of detail, even within the same resource.  For example, in the Reactome database, some reactions consider ERK1 and ERK2 independently, whereas others mention only a combined ERK1/ERK2 entity.  Unfortunately, the link between these species is not explicitly defined in the database.  Better semantic tagging of species and reactions in these resources could help alleviate these problems. 

Interestingly, these challenges are not unique to building a model from information in databases and online resources – modelers themselves face these same issues when building models from the primary literature.  To me, this demonstrates that although these resources may facilitate model construction, they are currently not in a position to revolutionize the careful manual construction of models.  Certainly, the fact that about half of the reactions and species in the EGFR pathway that the authors expected to find were not retrieved at all, even after additional database queries, means that literature review must remain a vital part of model building.  So despite the large amount of content in online pathway databases, the incompleteness and differing resolution of the data means that model building will likely continue as a manual process, at least for the foreseeable future.

Neuroscience and Systems Biology

By Vy Hoang

With recent advances in genetic and functional genomic methods, the field of neuroscience is playing catch up with the other areas of biology that have already adopted these techniques (Geschwind et al.). Some reasons behind its slow start are related to the complexity of the neural circuits, the dependency on post-mortem human tissue, and the massive amounts of data generated from neuroscience laboratories. However, these new technologies will also generate lots of data across multiple levels (phenotypic, genotypic, epigenetic, proteomic, and genomic). The question now is how do neuroscientists integrate these different layers of data? The answer is that they largely haven’t!

Using a reductionist view, neuroscientists have traditionally divided the brain into anatomical regions and have then tended to study each region in detail, but also in isolation (Villosalda et al.).  Although this approach has had its successes, the next step to understanding the complexity of the brain as a whole with regard to functions like memory, learning, and behavior is to integrate the knowledge of the different anatomical regions. Naturally such an ambitious goal is easier said than done.  Some sort of standardization in the protocols used for neuroscience experiments might be a step in the right direction, allowing researchers to combine experimental data more readily.

Neuroscientists might do well to consider the systems biology approach in which the components of a system are considered at multiple levels of detail whose properties are interrelated (Villosalda et al.). By enabling neuroscientists to integrate data in ways that allow them to study the brain as a whole, systems biology may help them to understand the mechanisms underlying neurological diseases such as Parkinson’s disease and schizophrenia since their etiologies involve more than one section of the brain. The elucidation of these disease mechanisms would enable medical researchers to more clearly define neurological targets of interest and to develop safer and more effective therapies. Although the field of neuroscience may currently be lagging behind in embracing the systems biology approach, it seems likely that its researchers will also seek to take advantage of the benefits that it can offer before too much longer.

The Death of Skepticism

By Isha Antani

In recent times, systems biology has been “rediscovered”. Systems biology has been around for a while, often referred to by different names. Papers were published as early as 1952 that explored the concepts of modeling cellular systems (e.g., Hodgkins and Huxley’s Nobel Prize winning paper on a mathematical model of the action potential propagating through an axon). Until recently, however, it was never much in the mainstream and scientists, for the most part, continued to focus more on specific molecular biology or on looking at biology from a larger, cellular perspective but never forging a link between the two. Since the turn of the 21st century, this area of science has been slowly gaining steam. The International Conference on Systems Biology (ICSB) was started in 2000 and each annual conference has been growing in number of attendees ever since. It was only a matter of time before the pharmaceutical industry recognized the potency of the field. However, as quickly as the pharma companies jumped on board, they hopped off board when the science failed to deliver fast results.  

Now given some room in which to breathe, the field of systems biology has gotten time to flesh out and develop. A multidisciplinary field, merging computer science, biology, and engineering, it has now had the needed time to create a strong message and a thriving community. In the last few years, technology catering towards this field has improved considerably and has led to the acquisition of and better digestion of large, meaningful data sets about protein-protein interactions. Additionally, as better-trained scientists emerge from the multitudes of recently created Systems and Computational Biology university programs, the field has started to generate tangible results from model-driven experiments.

For starters, Cambridge-based Merrimack Pharmaceuticals has successfully developed the first drug ever motivated by a systems biology model, MM-121. MM-121, an ErbB3 inhibitor, entered phase 1 in late 2008. In June 2009, another Merrimack drug MM-111, which simultaneously targets ErbB2 and ErbB3, entered clinical development in a Phase 1/2 trial. 

Pharmaceutical companies such as Roche that tentatively invested into the field have also started seeing effective results and quicker than expected. In 2005, Roche collaborated with SystemsX, a Swiss Initiative in Systems Biology, on a project titled Systems biology of the beta cell- application to type 2 diabetes progression. There was some skepticism about what this partnership would be able to accomplish in the three-year run the partnership was initially given. However, within a year of the project start, the group headed by Cristiano Migliorini was able to drill down to a single, clear target. The identity of the target protein is still kept secret. 

As additional systems biology-derived drugs enter the pipeline in the next few years and eventually make it into the marketplace, then perhaps in the next few years skepticism in the pharma industry will diminish. Surely enough, a recent report by PricewaterhouseCoopers stresses the need for the incorporation of in silico methodologies in lab regimens and predicts that “by 2020, virtual cells, organs, and animals will be widely employed in pharmaceutical research, reducing the need to experiment on living creatures.”

What's New In Cellucidate?

By Traci Freed

Exciting new functionality, including collaboration tools, has recently been implemented in Cellucidate.  Because Cellucidate is web-based there are no updates to install on your computer so login to your account and start exploring these features.  If you are not a registered user, you can sign up at www.cellucidate.com.

Collaboration Tools
Cellucidate has made it easy to collaborate with other users.  Just add users as collaborators to your cBook, choose which permissions to give them and you have an online workspace where you can share your work and draw on the contributions of other members in the Cellucidate community.  To learn more about sharing cBooks, go to our tutorial page and watch the third video. There is also a bookshelf dedicated for cBooks shared with you so you can quickly navigate to them and keep them organized.  And to effectively communicate with other users when working in Cellucidate, you now have the ability to send and receive messages to them through the system. 

Status Bar
When viewing a cBook there is a new status bar located at the bottom of the page that enables you to better manage your work.  With the status bar you can:

• See which collaborators are simultaneously viewing the cBook
• View the models and rules of a cBook by navigating to specific pages using the contact map
• See jobs running in the cBook
• View updates to your cBook
• Navigate to messages you have sent to and received from other users

Other recent improvements to Cellucidate include the ability to import Kappa files and download a snapshot of the information generated from a simulation in a CSV file.

We are always looking for ways to make the Cellucidate experience better for you.  Email your comments to feedback@cellucidate.com or add them to the Feature Requests topic under Support in the Discussions forum in Cellucidate.

Transparent Models

By Isha Antani and Ty Thomson

Understanding and interpreting a differential equation model published in a paper can require one to go beyond merely copying and running the model in a simulator. Sometimes, running someone else’s previously published model generates data that do not agree with the data published in the paper. A researcher who encounters this issue has to go back to the original model and pick apart each and every term in the differential equations to make sure that the term was correctly copied or translated.

Because the original model is often just a text description, and not an executable file, it can be difficult for the researcher to assess whether differences between their simulation results and the published results are due to an error of their own, or due to an error in the original paper.
Unfortunately, errors in published models are not as uncommon as one might hope, perhaps because of the onerous task of re-implementing models during the peer review process.

Even when the original model is available in an electronic and executable format, it can be difficult to read through the model (which is often in the form of computer code) to discern why the two sets of results do not agree.  Perhaps there is something in the description of the original model that disagrees with the executable model.

For models to be easily replicable, transferable, and understandable, they have to be transparent. In this context, transparency means that the representation of the model has to be easy to read and understand, and should thus contain appropriate documentation.  For example, a viewer might be able to click on an equation in a transparent model to visualize the corresponding reactions and assumptions. Such a transparent system, furthermore, can aid the creator of the original model to easily flag any flaws that they might have introduced while creating the model.

There is an increasing demand for ease of sharing models of signal transduction pathways. As the study of pathways and signaling systems spreads, there are bound to be an increase in the number of papers that have model accompanying them. Therefore, having these models be transparent and easily reproducible is going to become even more of a necessity in the future.

10th Annual International Conference on Systems Biology

By Joshua Havumaki and Ty Thomson

Many from Plectix attended the International Conference on Systems Biology (ICSB) last week at Stanford University.  In attendance were hundreds of attendees from all over the world.  Although the vast majority of participants were academics, the commercial sector was also represented.

The talks were quite diverse and ranged from the usual, high throughput methodologies to light sheet based fluorescence microscopy. However, much of the work presented was centered on specific pathways of clinical relevance as opposed to general tools or methodologies. Talks at the q-bio conference in early August were a bit broader in scope, describing tools and methodologies (both experimental and computational) that can be applied to the system or cellular level. Despite the differences in the talks, the attendees at both conferences seemed to have similar backgrounds judging by the posters presented.

The tutorials at ICSB were well attended and featured many tools for deterministic and stochastic modeling, including Virtual Cell and Systems Biology Workbench. In our own experiences (both attending tutorials and demoing Cellucidate) there is a definite necessity to provide more illustrative and interesting examples and a more interactive and hands-on approach to tutorials leads to a more successful outcome of engaging audiences. We certainly have struggled with this problem and there is more to be done in this area.

We presented two posters that covered some representative models in Cellucidate and the computational approaches behind the scenes - "Modeling Dysregulation of Complex Systems Implicated in Cancer and Diabetes" and "Scalable Methods for Modeling Complex Cell Signaling Systems". Both generated a steady stream of interest during their respective poster sessions. Ty Thomson also gave a polished, though somewhat impromptu talk on rule-based modeling and Cellucidate at the 14th Annual SBML Workshop.

It would have been nice to have more time set aside for attendees to discuss different perspectives and opinions of the future in the field. However, overall this conference did provide a comprehensive picture of current developments and research in systems biology.