Systems Biology: A Very Short Introduction

Very Short Introductions #634

by Eberhard O. Voit

Systems biology came about as growing numbers of engineers and scientists from other fields created algorithms which supported the analysis of biological data in incredible quantities. Whereas biologists of the past had been forced to study one item or aspect at a time, due to technical and biological limitations, it suddenly became possible to study biological phenomena within their natural contexts. This interdisciplinary field offers a holistic approach to interpreting these processes, and has been responsible for some of the most important developments in the science of human health and environmental sustainability.

This Very Short Introduction outlines the exciting processes and possibilities in the new field of systems biology. Eberhard O. Voit describes how it enabled us to learn how intricately the expression of every gene is controlled, how signaling systems keep organisms running smoothly, and how complicated even the simplest cells are. He explores what this field is about, why it is needed, and how it will affect our understanding of life, particularly in the areas of personalized medicine, drug development, food and energy production, and sustainable stewardship of our environments. Throughout he considers how new tools are being provided from the fields of mathematics, computer science, engineering, physics, and chemistry to grasp the complexity of the countless interacting processes in cells which would overwhelm the cognitive and analytical capabilities of the human mind.


ABOUT THE
The Very Short Introductions series from Oxford University Press contains hundreds of titles in almost every subject area. These pocket-sized books are the perfect way to get ahead in a new subject quickly. Our expert authors combine facts, analysis, perspective, new ideas, and enthusiasm to make interesting and challenging topics highly readable.

Genres: Science, Biology, Audiobook, Nonfiction

160 pages, Paperback

Published June 1, 2020

Reviews

[James · Rating 3 out of 5]

Decent summary of what systems biology is all about, but had less new content that was new to me than I expected. Interesting general principles about complex systems and high-dimensional science, and some interesting specific tidbits. Potential inspirations here for how to do transparency research - all the gears are available, but the understanding is lacking. Author deserves some slack though, because this topic is technical, relatively new, and one of the central problems of the field is the lack of universal results.

Notes:
• Systems bio tries to understand biology in terms of modularity - organisms can be broken down into systems, which are interconnected in generalizable ways and composed of subsystems to which the same analysis can be applied.
• Modern computers, microscopy, and high-throughput research has let us catalogue microstructures of biology to an insane degree. We're getting close to understanding the full molecular inventory - having all the gears. Systems biology is about taking the gears and figuring out the model.
○ Take inflammation as an example: we have a system with known gears, a lot of known behavior, but still lots of poorly understood edge cases we don't know how to intervene in or prevent.
• Systems become harder to understand with size, nonlinearity, feedback, phase transitions, redundancy
○ A bunch of systems in series which have phase transitions seems likely to produce chaos
• A lot of unpredictability also comes from the variety of different regulatory control mechanisms, which act in different ways and on different time scales
○ Feedback on enzymes: to regulate the amount of X, things are arranged so that X down-regulates one of the steps in the enzymatic pathway leading to its production. Downregulate a producing enzyme, or upregulate a competitor, or upregulate an enzyme that degrades a producing enzyme, or something like that.
○ Feedback on transcription/translation: to regulate X, X down-regulates one of the steps in the pathway leading to the production of the enzymes leading to its production. Reduce gene expression via epigenetics or inhibit protein translation somehow
• There's also forcing inputs that have to do with the environment rather than the organism - trehalose protects yeast against high temperatures, and the enzymes that produce it have their highest efficiencies above yeast's optimal temperature
• Different speed pathways can be used to "program" certain response curves: if something triggers a short-term up-regulator and a long-term down-regulator, the variable in question will go up first, then down after a set amount of time
• In the era of massive data, scientists have moved away from formulating and testing specific hypotheses, towards massive exploratory data analyses.
○ This is johnswentworth's "science in a high-dimensional world" thing.
○ Fits well with the idea of systems biology - when the question isn't "what exists" but "what does what", massive search is needed
• Computational Systems Biology (CSB) uses ML, and also explicit models which can be either static network models, dynamic network models, or differential equations
○ Static networks can be evaluated very efficiently using matrix representation and linear algebra
○ Dynamic networks are much harder to solve, so they tend to be used for smaller systems where more accuracy is needed. Have "flux" arrows and "signal" arrows - all arrows start from nodes, but flux arrows influence the values of other nodes, while signal arrows influence the effects of other arrows
○ Differential equation models are sort of the opposite of high-dimensional science. They give you really powerful results, but you have to be confident that you've identified the relevant variables and how they interact
• The whole process of systems biology is all about hypothesis formulation - simulated or modelled results always have to be sent to lab to be validated
• Many biological networks have the "small-world property" - their average node separations are much lower than would be expected of a random graph with the same average degree, because some nodes are highly connected "hubs" and others are less connected.
• 2% of DNA is protein-coding
• Genomics focuses on forced feedback networks mediated by environmental sensors and transcription factors. Genes don't interact directly, but regulation of genes is performed by proteins from other genes, and proteins can interact with proteins directly.
• Proteomics is split into three main questions - availability, structure, and function. Roughly: "what can we learn from protein concentrations in different locations?", "how do we predict protein structures from gene sequences?", and "how do we infer function from structure?"
• Multi-scale modeling is extremely difficult - when we want predictions at the macro-scale, we can't really simulate all the way down to the lowest level of resolution. But sometimes small changes at the lowest level can trigger macro-scale changes
○ One approach is modularization: find ways to break the multi-scale system down into subsystems, then try to understand each of those subsystems. If we can find computationally efficient approximations for their behavior, then we're back in business
○ Question is, how to partition the system into modules?
• Analytical chemistry can produce biological compounds, but will tend to produce an equal mix of both chirality versions. Sometimes, reversed-chirality compounds can be toxic - and because of their highly similar properties, they are often difficult to synthesize. Biological organisms, however, produce all one chirality.
• Three approaches to metabolic engineering:
○ Flux balance analysis, start with a network model and perform constrained optimization on flux strengths to maximize output while letting organism survive
§ How is this implemented? You can't just step in and "crank up" the enzymes responsible for the fluxes… needs to eventually recurse back to environmental factors we can control with reactor design, or genetic factors we can alter
○ Elementary mode analysis, which identifies pathways responsible for functions that are no longer necessary under bio-reactor conditions, and eliminates them to free up resources
○ Brute force ODE modeling and constrained optimization. Come up with a huge model that's as detailed as you're able to do, and let a computer solve it
• Plants tend to have large numbers of chromosomes and large numbers of genes, making genetic modification complicated
• Biology does have laws of sorts: the genetic code, the law of competitive exclusion, etc. They do have exceptions, but so do a lot of other really useful "laws" in physics and chemistry.
• Complicated inter-connected systems lead to lots of necessary causes, but very few sufficient causes.

[Bill Tyne · Rating 2 out of 5]

To say this is a shallow over view of a much hyped subject is to underplay the nature of this volume. We may be shown the modeled puddle that is the much vaunted world of systems biology but nary a toe is dipped into its supposed depths.
What is systems biology - think a pinch of mathematical biology, bioinformatics and a modellers wet dream. Yep nothing new here other than the label and, frankly the rate the labels get reinvented and changed makes even the new name old hat!

[Nilendu Misra · Rating 4 out of 5]

An excellent walkthrough of how computational thinking, coupled with our ability to process gigormous data rapidly, is turning hypothesis-driven scientific method upside down. In the twentieth century, you would start with a hypothesis and then conduct specific experiments, hopefully, to falsify it. Today, you gather bunch of data, create “features” (what are you interested in), borrow a computational method (e.g., logistic function), or an emsemble of it, and then run it through massive, or many, computers. What you find would not be what you were looking for. But it could be more important than anyone would have thought on their own. Especially on truly complex domains like biology this is and will continue to create revolution. Our goal as a species has long ceased to answer “what all is there”. We are now fine-tuning “how the heck..” (e.g., design a ligand that could bind to a soecific xell receptor without messing things up). Delightful little book!

[Anna · Rating 4 out of 5]

Good first introduction to computational systems biology. The book would have benefited from footnotes supporting specific claims. A small bibliography attends the back, for further reading.

[The · Rating 5 out of 5]

Too technical

[Nadvornix · Rating 3 out of 5]

My rating is probably influenced by my attempt to listed to audio version. But I kept getting lost. This book do not translate to audio very well.

[Marco · Rating 3 out of 5]

An accessible introduction to technical aspects and significant breakthroughs, accompanied by some pedestrian philosophy of science.