Using computers to design living systems sounds like something you might see in a science fiction film. But speak to the engineers and scientists working in the emerging field of synthetic biology, and the idea becomes more feasible. Ultimately, they hope to develop the tools that, at the push of a button, will enable researchers to design, model and create DNA that can be inserted into a bacterial cell. The new DNA will then transform the cell into a hub of industrial activity, churning out a biofuel, plastic or even a vaccine.
This goal is no fantasy, according to Professor Richard Kitney, professor of biomedical systems engineering in the department of bioengineering at Imperial College, London. “I believe that in the next five years or so, you will see fairly sophisticated software that allows high-level design of bacterial DNA and how the bacterial DNA responds as a cell,” he says.
Kitney and his colleagues are working on BioCAD software for synthetic biology that is the direct equivalent of the CAD software used in other areas of engineering. The software allows engineers to modify DNA and simulate how the design would function in a cell before any work is done in a “wet” laboratory. His team are working to integrate these tools into a web-based information system, so users can log into the program and use it over the internet.
Software tools are likely to become increasingly important in synthetic biology, as engineers create more complex systems that would be too difficult to design in the laboratory purely by trial and error.
Having the software tools to programme cells could change the biotechnology industry to the same extent that software for programming silicon changed the computer industry, believes Andrew Phillips, head of the bio computation group at Microsoft Research in Cambridge. “This could transform society, in areas ranging from health to energy and agriculture,” he says.
Phillips and his colleagues have been working on software to programme cells for four years. The group’s aim is to programme cells as readily as computer programmers can programme silicon. He says developments in computer programming have brought this goal within their grasp. Software developers can now write computer code without having to understand the details of low-level machine instructions. Programmers write code in a high-level language that a compiler translates into a sequence of 0s and 1s, that the machine then reads and executes. Software that programmes cells follows a similar path, but produces a DNA sequence instead of binary code.
Microsoft has developed prototype software that works exactly in this way. It allows scientists and engineers working in synthetic biology to design genetic circuits without writing the DNA code. They simply list the behaviour of the cell they want to create – for example, how they want the proteins to interact – and the software automatically selects the DNA needed.
This software is called Genetic Engineering of Cells (GEC), and is already being used by researchers and students in laboratories in the UK and US. One of the advantages of using tools such as GEC is that researchers can save time and resources by simulating and analysing a system virtually before creating it in the laboratory. Creating a system in the laboratory can take weeks, if not months, and is costly.
A limited number of DNA sequences are built into the GEC software, and this limits the complexity of the biological system that researchers can create. Although lots of DNA has been sequenced, much of it has not yet been characterised, so scientists do not fully understand how it behaves. Without this information, it is impossible to predict how these DNA sequences will interact with others when modelling.
There is still much work to do, but Phillips believes the key challenge in programming cells is creating the tools that can compile these sequences of DNA. Work in this area is “pushing the boundaries” of research in computer science and biology, he adds.
“Computer programs traditionally have been sequential, like a recipe, whereas biological programming is much more difficult. In the chemical soup of the cell, things are all happening at once, and could interfere with each other,” he says. This means that it is much more difficult to predict the behaviour of a concurrent biological system than that of a system in which events happen in sequence.
Phillips and his team are using advanced methods for handling this concurrency. Programs that are highly concurrent, known in the industry as multi-core or multi-threaded, are becoming increasingly common even in traditional computing, he says. “This is one of the major challenges facing the computer industry today.”
The environment inside a living cell brings added complications. Biological processes fail, proteins become denatured and cellular matter is degraded, yet the cell continues to function reliably. Traditional computer chips do not behave in this way, says Phillips. With chips, if one element fails, the whole system can crash, as there
is no assumption that things will break as it runs. Synthetic biology software must be able to deal with these characteristics.
These problems are also being tackled by Carlos Olguin, head of the bio/nano programmable matter group at design software firm Autodesk. Working with biology throws up challenges that software designers have not seen before, he says. “It works in a very different way. We are used to designing top-down all the time.
“This new design space is one where you set local constraints on the parts, let the parts interact, and from that some kind of emergent behaviour results. That is a dramatically different way of thinking. We want to build abstraction on top of those bottom-up approaches.”
Despite these differences, patterns can be drawn from design software used in other industries – such as manufacturing, automotive and aerospace – and applied to synthetic biology, says Olguin.
One such notion is “scan, modify, print”, or “digital prototyping”. This has proved a powerful tool in manufacturing and construction, allowing engineers to visualise and alter designs virtually before beginning physical work on a product. It is a “natural progression” for Autodesk to apply these concepts to biology, says Olguin. “We are convinced it’s going to be the most important design space in the years to come,” he says.
Further down the line, Olguin sees a future where medicine is personalised via the scan, modify, print pattern. For example, a patient’s blood could be scanned at the point of care, medicine designed digitally based on the information gleaned, and then printed out for the patient.
Autodesk’s work so far in synthetic biology has included a partnership with the International Genetically Engineered Machine (iGEM) competition. The competition sees teams of undergraduates from universities worldwide competing by using synthetic biology techniques to design and build a biological system that can operate in a living cell to solve a specific problem. Competitors are using the company’s 3D animation software, Maya, to help visualise their entries and, in some cases, perform qualitative simulation and create plug-ins.
Olguin and his colleagues are working to expand these abilities in a more focused way.
The impact that software will have in the life sciences is giving rise to a lot of spin, says Professor David Gavaghan, professor of computational biology in the department of computer science at the University of Oxford. Scientists will come to depend on computer modelling more and more as biology becomes more qualitative but this could take decades, he says.
Gavaghan’s own work involves creating computer models that simulate the development of colorectal cancer. Modelling DNA using computer software is more straightforward than modelling cancer development, he says, because DNA as a system is simpler and easier to understand.
Creating software that can model, design and predict the behaviour of DNA is “realisable” today, he adds. Phillips’ work at Microsoft is “among the best” Gavaghan has seen in terms of translating computational work into prediction and putting that into experiments.
What is synthetic biology?
Synthetic biology is an emerging science that promises to revolutionise industry. It sees scientists and engineers design and build biological systems that can perform specific functions. Researchers create sequences of synthetic DNA that are then inserted into bacterial cells, programming them to produce substances such as plastics, fuels, or drugs. Essentially, synthetic biology is the “engineering of life”.