While computational systems biology often tends to focus on technical challenges such as integrative modeling and network reconstruction, the field is rapidly beginning to put those methods to use in applications targeted toward human health, according to speakers at a conference in New York this week.
There has “been a tremendous shift towards using regulatory, integrative genomic, and kinetic models to understand the biology of drug development [and] disease and using modeling and simulation to … predict what the cell is doing,” Andrea Califano, a professor of systems biology at Columbia University told BioInform during the 3rd Annual Joint Conference on Systems Biology, Regulatory Genomics, and Reverse Engineering Challenges held at Columbia this week.
Califano added that this shift means that the community is finally able “to put its money where its mouth is” and show that its methods and approaches can be used to make useful predictions about cellular systems that promote a deeper understanding of disease, as well as to identify potential drug targets.
His point was well reflected by the variety of research projects presented during the conference's two-day systems biology track, many of which used computational approaches to unravel the mechanisms underlying human biology and disease.
Cancer's Complexity
Cancer-based research occupied a significant share of the talks at the conference, with many presentations addressing the complex biology of cancer as well as methods to link genetic features of the disease to treatments.
A recurring theme underlying many of these projects was the complex nature of the disease and the fact that there are multiple external factors that influence its progression, such as the cellular microenvironment, genetic mutations, and other potential perturbations that have made unraveling the disease mechanism quite challenging.
However, Califano noted that cancer research has several advantages over systems-based research into other disease types, such as schizophrenia, autoimmune disease, and obesity, in which the associated cell types and pathways aren’t as clear cut. In the case of cancer, he noted, researchers have an advantage because cell types are often “reasonably well defined.”
Califano said that systems biology could impact cancer research by shedding light on “the mechanisms of addiction,” which describe how some genetic abnormalities in cancer cells depend on certain genes to keep the cells alive. In addition, he said that systems-based approaches could help researchers understand drug resistance and drug sensitivity in cancer patients.
For example, in one presentation, Chris Sander from Memorial Sloan Kettering described the use of a partial least squares regression approach to study the “interplay” between tumor cells and cells in the cancer’s microenvironment that may impact its progression, invasion, and metastasis.
During his presentation, Sander described a statistical method developed in his lab to create predictive network models of cancer cellular systems that have been perturbed by factors such as drug treatments and genetic alterations.
Other projects use computational approaches to reconstruct regulatory networks that are linked to cancer. For example, a project from a team in Califano’s lab at Columbia aims to use reverse-engineering algorithms to reconstruct regulatory programs in mouse and human prostate cancer models to create molecular interaction maps that could be used to identify master regulators, as well as possible drug targets for cancer.
Another project, a partnership between Columbia and Harvard University, used a probabilistic dynamic approach to simulate kinase signaling in a known cancer pathway.
'Organizing Data'
While Peter Sorger of Harvard Medical School agreed that that there is increasing interest in the systems biology community to answer more practical questions related to human biology and disease, he underscored the importance of adding more data to the mix in order to improve researchers' understanding of these systems. In particular, he stressed that it will be important to gain a better understanding of cellular phenotypes, because “it’s very hard to understand what networks mean if you don’t relate them to phenotypes or functions.”
Speaking with BioInform after his talk, Sorger noted that there are two aspects of systems biology: one that is "well developed" and relies on information such as genomic sequence and ChIP-sequencing data; and another, more “nascent,” aspect to the field that deals with other kinds of information such as advanced imaging, localization dynamics, and biochemical data.
Currently, “a lot of the activity” in this space lies in “trying to work out how you could organize and analyze data in such a way that you can make the kind of quantitative inferences that you can and need of data,” Sorger said.
He added that a further challenge lies in organizing the data in both areas and linking them together, “because really we are not going to understand how those biological processes work unless we can get both the protein dynamics worked out and also the genome dynamics.”
Sorger said that some challenges that the community needs to address going forward include “working out how spatial temporal dynamic data can really be quantified and shared.”
In addition, as interest in translational research grows, he said that new questions will arise with regard to whether the community’s efforts need to be geared toward using cells, mice, or human models — or, as he suggested, a combination of all three — to find answers.
Have topics you'd like to see covered in BioInform? Contact the editor at uthomas [at] genomeweb [.] com.