Musings on Complex Systems and Molecular Biology
Ever since my undergraduate in the University of the Philippines Diliman, the subject of Complex Systems has always piqued my interest. In our discussions of biological processes, such as the central dogma of molecular biology, cell division, metabolic pathways, and cancer generation, complex systems and the theories of complexity science have always been in the background.
We first discussed the concepts and processes in isolation, but towards the end, the idea that these processes connect to a bigger picture — or to a more “complex system” — has always been alluded to. Unfortunately, we did not go deeper than that as no one in the National Institute of Molecular Biology and Biotechnology has the expertise to formally discuss complex systems. This left the feeling that I was missing out on a bigger picture, that while I can study these processes in isolation, we were not really looking at it from a systems perspective. And it turns out that that gut feeling was right. In doing this essay, I had to research what is complexity science and what are complex systems. I also looked at how complex systems can interplay with the things we discussed in undergrad, and this led me to sources that really put into context what I missed out on in undergrad.
MacDonald, et al., (2012) describe complexity science “as the study of a system using not a single theory but a collection of theories and conceptual tools from an array of disciplines.” For a more formal definition, “complexity science is concerned with complex systems and problems that are dynamic, unpredictable, and multi-dimensional, consisting of a collection of interconnected relationships and parts.” Also, unlike linear thinking, where “cause and effect” is the paradigm, complexity science is characterized by non-linearity. MacDonald also stated that “in dealing with complex systems, we need to appreciate the dynamic and interconnected relationships occurring within the system or the problem.” One good example she provided was solving the problem of obesity, where there was a multitude of factors that contribute to the problem. Instead of reducing problems to their smaller parts, public health professionals would be more effective if they understood the complex relationships occurring between the factors.
Complexity in Biological Problems
Uthamacumaran (2020) also gave a similar definition of what a complex system is. More specifically, a complex system is a nonlinear dynamical system of many interacting parts which adaptively respond to the perturbations of their environment. The signatures of a complex adaptive system include nonlinearity, emergence, self-organized patterns, interconnected multi-level structures, critical phase-transitions, computational irreducibility, unpredictability and multi-scaled, feedback loops. In simple terms, the concerted whole cannot be defined by the sum of its interacting parts. Complex systems are also chaotic — moving or modifying small sections or pieces can have unexpected consequences. Unlike MacDonald though, Uthamacumaran gave concrete and more in-depth examples of how complexity science is applied to biological problems. For example, the reconstruction of cancer networks from gene expression data was discussed as a computational complexity problem. He also tackled the development of algorithms applied in stem cell reprogramming, i.e. how these algorithms can chemically alter the epigenetic landscape of differentiated cell states and how they differ from each other. There were also discussions on how Machine Learning algorithms are used in reconstructing cancer gene regulatory networks as well as protein-protein interactions that can help elucidate cancer network dynamics and identify the impact of varying conditions.
Admittedly, when I was reading through Uthamacumaran’s paper, I felt overwhelmed as there were a lot of things that looked foreign to me. On a very high level, I understood the biological processes they were discussing and what they were trying to achieve, but not the specifics and the technicals of the processes in between. At the same time though, it made me excited. Excited because I feel like the Network Science course of the Asian Institute of Management’s MS in Data Science program will finally give me a solid foundation to go through these papers and understand them a little bit more, similar to how the past Machine Learning 1 and Machine Learning 2 courses gave me the foundation of going through machine learning and AI-related papers.
In hindsight, it was bliss not really knowing what complex systems are, as we could always tackle the concepts in isolation and avoid the mathematical rigor of dealing with complex systems, but now that it is right in front of me, I can’t help but feel excited that I am FINALLY going to get that missing piece of the puzzle.
References
MacDonald, M., Jackson, B., Best, A., Bruce, E., Carroll, S., Hancock, T., . . . Riley, B. (2012). The relevance of complexity concepts and systems thinking to public and population health intervention research: A metanarrative synthesis. CIHR Knowledge Synthesis Grant. Retrieved from https://www.uvic.ca/research/groups/cphfri/assets/docs/Complexity_Science_in_Brief.pdf
Uthamacumaran, A. (2020). A Review of Complex Systems Approaches to Cancer Networks. Complex Systems, 779–835. doi:10.25088/complexsystems.29.4.779
