Feature - A SLiM chance for viruses

Feature - A SLiM chance for viruses

Viruses hijack the replication machinery of cells. Image courtesy Simon Hettrick

Viruses have evolved a clever way of reproducing. They hijack the replication machinery of their host cell, which is controlled and regulated by a variety of signaling pathways, and fool them into producing copies of the virus.

Richard Edwards, head of the Bioinformatics and Molecular Evolution group at the University of Southampton, UK, is trying to better understand signaling pathways in order to develop treatments for viruses - and for diseases that operate similarly. This is an enormous task, because to understand signaling pathways in the human body requires studying the interactions between the 20,000 or so proteins contained within the cells.

To do so, Edwards is focusing on short, linear motifs known as SLiMs. "A protein can be thought of as a sequence of amino acids, like beads on a string" explains Edwards. "[SLiMs] consist of about three-to-five specific amino acids in the protein."

Because SLiMs control the ways in which proteins interact, they could be responsible for many signaling pathways.

They are potentially useful to viruses too. Because SLiMs are small, it's relatively easy for a virus to evolve a structure that mimics them and hijack the signaling pathway controlled by the SLiM. For example, Foot-and-Mouth-disease infects host cells by exploiting a SLiM called RGD. The structural proteins of the HIV and Ebola viruses contain PTAP, a SLiM which recruits a human protein to cause a virus to "bud," or acquire a fragment of the host cell to become part of the virus's external envelope.

It is likely that further unknown SLiM interactions remain, which may offer potential sites for therapeutic intervention. It's not just viruses; a greater knowledge of SLiMs could be vital to treat a variety of diseases caused by problems with signaling pathways, such as heart disease.

Edwards shows us SLiMFinder. Image courtesy Simon Hettrick


To investigate, Edwards wrote the SLiMFinder, a computer program that determines whether the properties of a protein are caused by SLiMs. But the potential was limited by the time it took for results to be generated.

Also, the early version of the SLiMFinder's interface required scientists to learn complex commands to use it. The interface may be familiar to Unix users (and anyone who used a PC in the 70s or 80s), but it is something that many scientists in this point-and-click age find difficult to use. "It's one of those Catch-22 situations" says Edwards. "The kind of people who are able to download and run (the early version of SLiMFinder) are the kind of people who are likely to develop their own software."

To make SLiMFinder more powerful and user-friendly, Edwards teamed up with OMII-UK, an organization devoted to developing free, open-source software and helping researchers benefit from the technology.

A workflow for SLiMFinder was developed using OMII-UK Taverna software, allowing repetitive tasks (such as data collection from databases and data manipulation) to be automated, resulting in substantial time-savings.

By developing SLiMFinder into a Taverna workflow, the software gained a user-friendly interface. Rather than having to learn and type commands, scientists can now use a point-and-click interface - an important advance, because it opens up SLiMFinder to a much wider audience.

Using Taverna also allows access to Taverna's myExperiment website, described by New Scientist as "MySpace for scientists," which gives scientists a place to share Taverna workflows, meet like-minded researchers and offer each other help.

SLiMFinder has also been adapted to execute on distributed resources like the clusters at the University of Southampton, which has significantly reduced its execution time. At this point, the developers have not quantified the improvement, but state it was a considerable increase in speed. Even better results are predicted when the software begins working on a larger cluster such as the UK's National Grid Service (NGS).

Richard believes that SLiMFinder is now user-friendly enough for any biologist to use, regardless of their computing experience. This, together with faster execution times, has made SLiMFinder a tool that could be very valuable throughout the biological sciences. As Edwards summarizes, "It's made the difference between a tool that people would think 'that'd be interesting, but I'm never going to use it' and something that people could try out to see if it's interesting."

-Simon Hettrick, OMII-UK. Edited by Seth Bell