Biomedical engineering research has advanced by leaps and bounds in recent years, but the field still falls short of most other types of engineering in one key respect: engineers’ ability to model their work.
With the enormous complexity involved in even the simplest of living organisms, developed over billions of years of evolution, true models tend to fall well short of the wide array of interactions taking place within an organism, making it much more difficult to rely on computers to expedite engineering research and development.
That could change with the announcement that a research team from Stanford University and the J. Craig Venter Institute have developed the first full computer model of the lifecycle of a living organism.
The researchers, led by assistant professor of bioengineering Markus Covert, tackled the problem of creating a workable computer model by examining an increasingly popular organism in bioengineering, Mycoplasma genitalium.
This tiny parasitic bacteria, which lives in genital and respiratory tracts, gained some notoriety four years ago when the Venter Institute successfully created a synthetic version of the organism that was able to be successfully cloned.
The bacteria was attractive for both projects because of its unusual simplicity, featuring the shortest genome for any independent organism at only 525 genes. By comparison, Escherichia coli – another popular bacteria in biology research – requires 4,288 genes and a human holds around 23,000 in all.
In addition to its comparatively simple genome, Mycoplasma genitalium is one of the physically smallest bacteria in the world.
Modeling a micro-organism
To create the model of their chosen organism, the Stanford team searched through more than 900 papers on the bacteria detailing every process throughout its lifecycle.
This data was readily available thanks to a strong trend in bioengineering to conduct experiments collecting data on the effects of activating or deactivating individual genes. This data was ultimately used to develop 1,900 parameters for cellular interactions, all of which were collected into 28 independent modules.
According to The Atlantic, actually running these modules to generate a full model of the organism required 128 computers running for up to 10 hours at a time.
But the purpose of the project is ultimately not about Mycoplasma genitalium, so much as the organism is a convenient test.
“The goal hasn’t only been to understand M. genitalium better,” said Jonathan Karr, a graduate student at Stanford and co-first author of the study. “It’s to understand biology generally.”
The advantage of this approach, however, is that running an integrated model allows researchers to examine the interactions of multiple genes, rather than just a single gene like the research used to develop the model.
“Many of the issues we’re interested in aren’t single-gene problems,” Covert explained. “They’re the complex result of hundreds or thousands of genes interacting.”
Whereas current research often provides new data without the ability to properly apply it, the Stanford team hopes to make it far easier to understand the practical implications of new data.
Catching up with CAD
The researchers explain that one strong limiting factor in the advancement of biotechnology has been the inability of engineers to virtually construct and test their ideas before actually going through the effort of creating the engineered organism.
Providing scientists with a model will allow them to create better supported hypotheses before getting to the point of actual experimentation and could reveal effects that might not appear with more standard manipulation.
“If you use a model to guide your experiments, you’re going to discover things faster. We’ve shown that time and time again,” said Covert.