Microsoft Program Adapted to Predict Stem Cell Outcome

John Lister's picture

Software developed to hunt down bugs in programs for Windows has been adapted for stem cell research. Though the two tasks sound nothing alike, they both involve calculating the link between cause and effect.

New Scientist reports that the software is known as the Reasoning Engine for Interaction Networks, or RE:IN. It's been made available through Microsoft's research unit to several scientists who specialize in stem cells. (Source:

A stem cell is effectively the blank slate of cells in mammals. As well as dividing to produce more stem cells (meaning that the body grows), it can develop into specialist cells that carry out particular tasks. This adaptability means the body can also use stem cells as a building material to fix damaged tissue. While scientists know a lot about stem cells, they are still trying to figure out how a stem cell decides what type of specialist cell to turn into.

Stem Cell Development Works Like Computer Process

Using RE:IN as a way to tackle this problem involved taking advantage of its use of a mathematical process known as formal verification. In simple terms, it tests that the input and output of a computing process match up in the expected way.

When Microsoft uses RE:IN it does so to test software for bugs. It's similar in principle to getting users to operate the software and check everything works as expected, but formal verification uses mathematics to make sure every possible variant is checked.

In normal use, RE:IN checks that an input creates the expected output; in other words the computer user does something and the software responds in the expected way. However, the verification process can be reversed to take an output and figure out the input; in other words, by looking at the software's outcome, it's possible to find out what instruction led to the outcome.

Research Reveals Cells Not As Complex As Thought

The stem cell researchers were able to adapt RE:IN so that rather than look at software inputs and outputs, it looked at the relevant data about stem cells before and after they developed into specialist cells. The lessons learned from this testing meant the researchers could predict with 70 percent accuracy how a particular cell would respond to a genetic change. (Source:

The findings suggest that the process of stem cells changing is much simpler than previously assumed. That means it could be easier for scientists to artificially control how stem cells develop, in turn making it more viable to artificially create specialist cells, and possibly entire organs.

What's Your Opinion?

Are you surprised that software checking tools can be adapted so neatly for biology research? Should Microsoft be involved in medical research, even indirectly? Can you think of other ways RE:IN and formal verification could be adapted to solve non-computing problems?

Rate this article: 
Average: 5 (10 votes)