Molecular when databases align, old drugs find new purpose

A computer program that acts as a drug+disease matchmaker can predict how drugs intended for one disease might actually be used to treat another, such as lung cancer and Crohn's disease.

Medicine, meet malady.

Using freely available databases, scientists are finding new uses for approved drugs. In fact, ones that treat ulcers and seizures might be repurposed lung cancer and inflammatory bowel disease, respectively.

It takes an average of 15 years and about $1 billion to bring a single new drug to market.

Repositioning’ drugs that have already been approved could allow developers to bypass rigorous tests required by regulatory agencies. This shortens the time and lowers the cost of getting a drug from the lab to the local doctor’s office. It’ll be especially useful for deadly conditions in need of better therapies.

Finding new uses for drugs happen often by chance. Excess hair growth was a side effect of chest pain drug minoxidil, now mainly used for hair growth in Rogaine. And then there’s Viagra, which was originally developed to treat hypertension and chest pain.

But why leave it up to chance? If you know that a drug switches genes on and off, and you know what happens to those genes in a disease, then that drug should be able to help treat the disease. Simple enough. In two studies published last week, scientists show that they can predict these such compatibilities.

Atul Butte and colleagues from Stanford developed a technique to match drugs with diseases – using an ‘opposites attract’ algorithm applied to publically available databases. New Scientist explains:

Drugs that have an opposite effect on gene activity to a particular disease could be good candidates for treating the condition. So the researchers devised algorithms to look for drugs that ramp up the activity of genes that are unusually quiet in tissues affected by a particular disease, and suppress those that are hyperactive in that disease.

The National Institutes of Health’s Gene Expression Omnibus contains the activity of thousands of genes; the Connectivity Map show patterns of gene activity change when exposed to drugs.

  1. The team mashed up the datasets and developed a similarity score for the gene signatures of every pair of drug+disease they compared (100 diseases, 164 drugs).
  2. Then they tested a few of these drug+disease compatibility predictions in the rodent models.

And as the Stanford news release explains: “sparks flew” and “therapeutic futures” were envisioned:

  • Anti-ulcer drug cimetidine (GlaxoSmithKline’s Tagamet), predicted to fight lung cancer, slowed tumor growth in mice.
  • Anti-epileptic drug topiramate (Johnson & Johnson’s Topamax), predicted to improve inflammatory bowel disease like Crohn’s disease, reduced damage to colon tissue in rats.

Could cheap, over-the-counter meds turn out to be the cures for deadly diseases? Clinical trials are needed, but the unknown benefits of many more drugs are just waiting to be discovered, Butte says.

One stumbling block according to WSJ is that some compounds are already off-patent, diminishing the financial incentive for a company to pursue it.

The lung cancer findings and IBD results were published as two papers in Science Translational Medicine last week.

Image by ebertek via Flickr

This post was originally published on