Bioinformatics versus Computational Biology

A hospital outbreak, the NIH definitions, the first models, and the skills each field needs.

Watch on YouTube

Chapters

  1. 0:17 A hospital outbreak, studied two ways
  2. 1:30 The NIH definitions
  3. 2:26 The first models in computational biology
  4. 3:20 What each role builds
  5. 4:10 Worked example: the SIR model
  6. 4:43 The skills each field needs
  7. 5:12 Summary

Transcript

0:17 A hospital outbreak, studied two ways

First, we will look at a hospital outbreak and at how each field studies the same virus. During an outbreak, a hospital laboratory sequences the virus genome from each infected patient. A bioinformatician aligns those sequences to a reference genome and compares the genomes with one another. The differences between the genomes place the patients on a phylogenetic tree, and the tree shows which infections likely share a source. A computational biologist uses the same outbreak to build a mathematical model. Equations describe how people move from susceptible to infected to recovered, and the model predicts how many new cases the hospital will see next week. In this outbreak, the bioinformatician works with the sequence data and the software that analyzes those sequences, and the computational biologist writes equations that predict how the infection will spread.

1:30 The NIH definitions

Researchers often use the terms bioinformatics and computational biology as if they had the same meaning. This video sets computational biology beside bioinformatics. In July 2000, a committee at the US National Institutes of Health wrote working definitions of both fields. The committee's definition of bioinformatics covers the research, development or application of computational tools and approaches for expanding the use of biological, medical, behavioral or health data. The definition of computational biology covers the development and application of data-analytical and theoretical methods, mathematical modeling and computational simulation techniques to the study of biological, behavioral and social systems. The two definitions come from the same committee: bioinformatics centers on tools and data, and computational biology centers on the study of biological systems.

2:26 The first models in computational biology

The difference shows most clearly in where computational biology began, with mathematical models. An early model of this kind comes from 1952. Alan Hodgkin and Andrew Huxley described the nerve impulse in the squid giant axon with four differential equations. Huxley solved those equations numerically on a hand-operated calculator, over about three weeks. The model predicted the shape and the speed of the action potential from measured ion currents. Hodgkin and Huxley shared the 1963 Nobel Prize in Physiology or Medicine with John Eccles. A second early model from 1952 came from Alan Turing, who published a reaction-diffusion model of pattern formation. Turing's model showed how two interacting chemicals can produce spots and stripes, and he proposed it as a mechanism for pattern formation in the embryo.

3:20 What each role builds

The two roles differ in what the person builds. A bioinformatician mostly builds or runs tools that turn data into results. Aligning sequencing reads, calling variants, searching databases and building pipelines are bioinformatics work. A computational biologist mostly builds models that explain or predict how a biological system behaves. Those models take the form of equations, simulations or statistical models. The research question decides which label applies. 'Which genes differ between these samples?' is a bioinformatics task. 'Why does this population of cells oscillate?' is a computational biology question. The two fields overlap heavily, and many researchers work in both fields. Job titles and degree names often use the two terms interchangeably.

4:10 Worked example: the SIR model

A worked example shows the modeling side, using the SIR model of an epidemic. It divides a population into susceptible people, infected people and recovered people, and it describes how people move between the three groups with two rates. The transmission rate is written beta, and the recovery rate is written gamma. The SIR model belongs to computational biology, because it predicts how the system behaves from a few assumptions. You estimate beta and gamma from case data, and that step is data analysis.

4:43 The skills each field needs

The two fields share a common base. Both fields need programming, usually in Python or R, and both need statistics. Bioinformatics leans towards command-line tools, file formats such as FASTA and BAM, databases and reproducible pipelines. Computational biology leans towards mathematics: differential equations, probability, simulation and optimization.

5:12 Summary

The shared skills complete the comparison between the two fields. In the definitions that a committee at the National Institutes of Health wrote in 2000, bioinformatics centers on tools and data, and computational biology centers on the study of biological systems. A bioinformatician mostly builds or runs tools that turn data into results, and a computational biologist mostly builds models, such as the SIR model, that explain or predict how a system behaves. The two fields overlap heavily, and both rely on programming and statistics.

Read more

All videos