The Systems Biology Vision of Dr. Mohammad Mobashir
The cancer cell is not a cell with a broken part — it is a cell whose regulatory architecture has been sufficiently perturbed that a new, self-sustaining dynamic has emerged. Understanding that dynamic requires a fundamentally different scientific vocabulary than the one built on isolated genes and linear pathways.
When a targeted therapy fails, the instinct is often to look harder at the target. Was the mutation misidentified? Was the binding affinity insufficient? These are reasonable questions, but they rest on an assumption worth interrogating: that the disease is primarily a problem of one molecule, one gene, one pathway. Complex diseases — cancer being the most instructive example — do not behave like single-component failures. They behave like system-level reorganisations. The cancer cell is not a cell with a broken part; it is a cell whose regulatory architecture has been sufficiently perturbed that a new, self-sustaining dynamic has emerged. The reductionist tradition in molecular biology produced genuinely foundational insights — identifying oncogenes, characterisingtumour suppressor proteins, mapping the kinase cascades that regulate cell division — but reductionism operates most cleanly when the relationship between a component and a phenotype is direct and stable. In complex diseases, that relationship is almost never either. A mutation in a signalling kinase does not have a fixed phenotypic consequence; it has a context-dependent one, shaped by the activity states of every other element in the network surrounding it. The same alteration can drive proliferation in one cellular environment and trigger apoptosis in another. Studying the mutation in isolation cannot tell you which outcome will occur.
This is where network thinking becomes not merely useful but necessary. When we represent a biological system as a network — nodes as proteins or genes, edges as their physical or regulatory interactions — we gain access to structural properties that have no analogue at the individual component level. Connectivity patterns, feedback loops, hub architectures, and the relative fragility of different nodes under perturbation all become visible. Signal transduction networks are particularly informative here. The received picture of signalling — a tidy cascade from membrane receptor to transcriptional output — was always an idealisation. What more comprehensive network analyses reveal is a web of cross-pathway interactions, competing feedback circuits, and regulatory relationships that make the linear picture insufficient for predictive purposes. A drug that blocks a node in this web does not simply suppress its downstream targets; it alters the dynamic equilibrium of the entire connected system. The network adapts. Alternative routes activate. Compensatory mechanisms engage. What we call treatment resistance is, at the mechanistic level, a description of this adaptation — the disease system finding a new stable state that bypasses the therapeutic intervention. Modelling this behaviour computationally requires both mathematical rigour and biological specificity: ordinary differential equations describe interaction kinetics; graph-theoretic analyses characterise structural vulnerabilities; Boolean and probabilistic models simulate how perturbations propagate through regulatory logic. The model is always a simplification, and knowing what has been simplified is as important as knowing what has been included.
“We are not short of biological information. We are short of the integrative frameworks required to interpret it honestly, and the intellectual culture required to resist the temptation of premature certainty.”
Multi-omics integration has added a further dimension to this work that is both exciting and methodologically demanding. Genomic, transcriptomic, proteomic, and metabolomic datasets each offer a partial cross-section of the biological system. The regulatory logic that connects them — how a somatic mutation reshapes transcriptional programmes, how those programmes translate into altered protein activity, how metabolic reprogramming follows — is precisely the kind of multi-layered information that systems-level models are built to handle. But coherent integration across these data modalities requires careful attention to how different measurement technologies introduce different sources of noise and bias, and demands analytical frameworks designed from the beginning to reason across levels rather than within them. The computational infrastructure for this is still maturing: we are past the point of naïve optimism about what data volume alone can achieve, but not yet at the point of having standard, validated methods for every class of integration problem. The interdisciplinary character of this work is definitional, not incidental. A computational biologist who cannot reason about the biological plausibility of a model’s assumptions will produce technically impressive but scientifically hollow outputs. I have spent considerable time structuring mentorship environments where students are pushed toward fluency rather than specialisation — where a student trained in bioinformatics is expected to understand the experimental context of the data she analyses, and where a student with a wet-lab background is expected to develop genuine computational intuition. Looking forward, the promise of precision medicine will only be realised through systems-level thinking. The current clinical implementation — matching therapies to genomic alterations — is a beginning, not an endpoint. A patient’s tumour is not a mutation; it is a dynamical system whose future behaviour emerges from the interaction of genetics, epigenetics, microenvironmental pressures, and immune dynamics. Predicting that behaviour, guiding therapeutic sequencing, and anticipating resistance are inherently network problems that require network solutions. What I believe, and what the evidence increasingly supports, is that our generation of researchers carries a specific responsibility: to build the conceptual and computational frameworks through which the data now available to us can actually be understood. Not processed, not visualised, not summarised — understood. That distinction matters. The era of data and networks demands not just new tools but a more rigorous, and more humble, relationship with complexity itself. The integrative frameworks required to interpret biological information honestly, and the intellectual culture required to resist the temptation of premature certainty, are not technical problems. They are scientific ones — and they define the frontier on which systems biology now stands.
References
- Barabási, A. L., Gulbahce, N., &Loscalzo, J. (2011). Network medicine: a network-based approach to human disease. Nature Reviews Genetics, 12(1), 56–68.
- Hanahan, D., & Weinberg, R. A. (2011). Hallmarks of cancer: the next generation. Cell, 144(5), 646–674.
- Tyson, J. J., Chen, K. C., & Novak, B. (2003). Sniffers, buzzers, toggles and blinkers: dynamics of regulatory and signaling pathways in the cell. Current Opinion in Cell Biology, 15(2), 221–231.
- Hasin, Y., Seldin, M., &Lusis, A. (2017). Multi-omics approaches to disease. Genome Biology, 18(1), 83.
- Kitano, H. (2002). Systems biology: a brief overview. Science, 295(5560), 1662–1664.
- Mobashir, M. et al. (2012). Simulated evolution of signal transduction networks. PLoS ONE, 7(12), e50905.
About the Author
Dr. Mohammad Mobashiris a Ph.D. in Biochemistry (Immunology) from Otto-von-Guericke University, Germany, where his research focused on mathematical modelling and evolution of signal transduction pathways and networks. He is currently a Researcher at the Department of Biomedical Laboratory Science, NTNU (Norwegian University of Science and Technology), Norway, and has served as Assistant Professor at King Abdulaziz University, Saudi Arabia, and as Postdoctoral Researcher at Karolinska Institute, Sweden, and Otto-von-Guericke University, Germany. He has authored 49 research articles with a total impact factor of 135.151 and an H-index of 14, with research spanning computational systems biology, cancer biology, gene expression profiling, pathway modelling, immunology, and high-throughput data analysis.