Best Place To BuildHow Computational Microbiology drives disease research & treatment | Prof Karthik Raman | BP2B S2 E8
CHAPTERS
- 0:00 – 0:34
From “spherical cows” to reverse‑engineering life
The episode opens with the classic “assume a spherical cow” joke to illustrate how modelers simplify reality. Prof. Raman connects this to why biology often feels intimidating—and why systems thinking is needed to understand complex living systems.
- •“Spherical cow” as a metaphor for simplifying assumptions in modeling
- •How students often choose “math vs biology” as an either/or
- •Reverse engineering as the default approach in biology (vs building from known components)
- •Why complexity in cells demands new ways of thinking
- 0:34 – 2:01
Show setup: IIT Madras, the podcast, and introducing Prof. Karthik Raman
The host frames the series as conversations with builders at IIT Madras and introduces Prof. Raman’s role in Data Science & AI. The stage is set for a fast-paced primer on computational and systems approaches to biology.
- •Podcast premise: what people are building at IIT Madras
- •Prof. Raman’s affiliation: Department of Data Science & AI (formerly Biotech)
- •Host’s motivation: demystify terms and concepts in Prof. Raman’s work
- •Transition into foundational definitions
- 2:01 – 4:33
What “systems biology” means (and how it differs from bioinformatics)
Prof. Raman defines systems biology as an engineering-oriented way to understand biology holistically. Instead of focusing on single genes or proteins, it aims to model interactions and predict how perturbations ripple through a cell.
- •Bioinformatics vs computational biology vs systems biology (engineering intersection)
- •Why reductionism (one protein/pathway) isn’t enough for cell-level understanding
- •Goal: gain a manipulable, predictive understanding of cellular systems
- •Perturbation as a core method: change one part and observe consequences
- 4:33 – 5:13
Modeling 101: simplified representations that answer specific questions
The conversation zooms into the meaning of “modeling,” starting with familiar classroom models (ball-and-stick). A model is framed as a deliberate simplification that captures key behaviors to make systems understandable and predictable.
- •Models mimic select characteristics of real systems
- •Scale and abstraction: what a model captures vs ignores
- •Models are built to answer a question, not to replicate everything
- •Biological models range from high-level to highly detailed
- 5:13 – 8:06
Pandemic math: SIR models, R₀, and why COVID challenged assumptions
Using COVID spread as an example, Prof. Raman explains epidemiological modeling through SIR differential equations and R₀. He emphasizes iterative refinement: when reality deviates (e.g., asymptomatic spread), models must evolve.
- •Rate-of-spread modeling vs predicting new variants
- •R₀ and what it indicates in outbreak dynamics
- •SIR model basics: Susceptible–Infectious–Recovered compartments
- •DBTL mindset: add complexity only when simple models fail
- 8:06 – 10:01
Digital twins vs “spherical cows”: accuracy, usefulness, and ‘all models are wrong’
The discussion contrasts lightweight approximations with the ideal of digital twins—models that behave like real systems. Prof. Raman defends starting simple, anchored by the quote: “All models are wrong, but some are useful.”
- •Digital twin concept: a model that mirrors real-world behavior closely
- •Why starting with simplifications is still valuable
- •Analogy to space mission modeling (treating planets as point masses)
- •George Box quote and the practical role of approximations
- 10:01 – 14:41
Microbiology modeling in practice: TB drug targets and cell-level selectivity
Shifting from populations to single cells, Prof. Raman explains how modeling helps identify drug targets in pathogens like Mycobacterium tuberculosis. The challenge is selecting essential microbial functions while avoiding human side effects and collateral damage to beneficial microbes.
- •TB example: 4,000 genes—how to choose drug targets?
- •Genes → proteins → catalyzed reactions as the mechanistic backbone
- •Essentiality and cascading effects across cellular networks
- •Selectivity constraints: avoid targeting human proteins or helpful bacteria
- 14:41 – 19:59
Computational systems biology and the rise of omics (genomics to metabolomics)
Prof. Raman distinguishes experimental systems biology (measuring at scale) from computational systems biology (integrating parts into predictive models). He walks through omics layers and explains why data integration is needed to connect molecules to system behavior.
- •Experimental systems biology: measuring ‘the whole’ (omics)
- •Central dogma refresher: DNA → RNA → proteins
- •Genomics, transcriptomics, proteomics, metabolomics—what each measures
- •Putting ‘parts lists’ together into system-level models
- 19:59 – 28:57
Microbiomes across extremes: gut, deep-sea vents, and the ISS (and why it matters)
The episode expands from gut health to environmental and extreme microbiomes, arguing that exotic ecosystems can reveal novel pathways. These pathways can be harnessed for green manufacturing and metabolic engineering—turning microbes into producers of valuable compounds.
- •Microbiome as ecosystem; microbes’ role in health across the body
- •Extremophiles and ‘exotic’ metabolisms in harsh environments
- •Metabolic engineering: transferring pathways (e.g., artemisinin production)
- •ISS example: surfactant production (“space to your face”)
- 28:57 – 33:52
Networks everywhere: graphs, pathways, and ‘Google Maps’ inside a cell
Prof. Raman explains networks/graphs as the unifying computational language of his lab—from microbial interaction networks to metabolite-reaction maps. The core idea is to convert biology into navigable structures so algorithms can find routes to desired molecules and functions.
- •Network basics: nodes and edges (social network analogy)
- •Networks at multiple scales: microbes, metabolites, reactions, even atoms/bonds
- •Stoichiometric reaction catalogs as a foundation for modeling
- •Google Maps analogy: pathfinding to reroute metabolism toward a ‘molecule of interest’
- 33:52 – 40:06
IBSE, Genome India, and metagenomic surveillance (Chennai’s microbial signature)
Prof. Raman outlines IBSE’s origin and evolution into integrative biology and systems medicine, emphasizing interdisciplinary collaboration and national-scale projects. He describes Genome India and MetaSub-style studies that profile city microbiomes and enable future public-health surveillance.
- •IBSE’s founding vision: quantitative minds tackling biological problems using available data
- •Genome India: sequencing and analyzing ~10,000 Indian genomes
- •MetaSub and metagenomics: sampling environments to profile microbial signatures
- •Chennai subway/city swab study and implications for antimicrobial resistance monitoring
- 40:06 – 49:18
Gut microbiome realities: antibiotics, recovery times, probiotics, and marketing hype
The gut is framed less as a ‘second brain’ and more as a major metabolic organ that produces compounds humans can’t. The host and Prof. Raman discuss how antibiotics disrupt ecosystems, why one-size-fits-all probiotics are limited, and how modeling could enable personalized interventions.
- •Gut as metabolic organ; microbial production of useful metabolites (e.g., vitamins)
- •Antibiotics as ‘forest fires’ with recovery spanning months to years
- •Individual variation: organism differences vs functional similarity across guts
- •Skepticism about supplement marketing; need for systematic design and validation
- 49:18 – 1:01:26
Education and career arc: chemical engineering → computation → systems microbiology
Prof. Raman traces how process control and modeling in chemical engineering primed him for systems biology, and how computational science training shaped his research style. The conversation broadens into how biology education should become more quantitative and integrated with math/AI.
- •Process control parallels: controlling reactors vs controlling cells
- •From disliking school biology to embracing molecular, quantitative biology
- •Why curricula push a harmful ‘math vs biology’ choice
- •The Genome Project’s lesson: diseases are network dysfunctions, not single-gene issues
- 1:01:26 – 1:09:44
Teaching philosophy at IITM and what’s next: microbiome dynamics and coral probiotics
The closing section covers Prof. Raman’s approach to teaching across departments and motivating students. He then previews near-term lab goals: modeling microbiome dynamics, designing minimal functional communities, and applying ‘probiotics’ beyond humans—such as restoring coral health.
- •Student-friendly, high-expectation teaching style; cross-department learning benefits
- •Next 2–5 years: shift from characterization to controllable microbiome interventions
- •Tooling focus: algorithms for microbial interactions and community design
- •Environmental applications: Andaman coral microbiomes and ‘coral probiotics’
