Using artificial intelligence to develop and test new drugs

August 3, 2016

We are in the midst of a crisis in drug development. The average drug costs approximately $5 billion and takes 12 years to go from pre-clinical development to market.

According to the U.S. Food and Drug Administration, “The drug research process is complicated, time-consuming, and costly…only 5 in 5,000 compounds that enter preclinical testing make it to human testing, and only 1 of those 5 may be safe and effective enough to reach pharmacy shelves.”

The bottom line: Some 95% of drugs fail to be deemed safe or effective for human use due to misleading test results from animal studies.


That’s why a team of innovative scientists at Insilico Medicine in Baltimore is utilizing the powerful capabilities of deep learning – often referred to as artificial intelligence – to open the door to a new era of drug development without animals, one that can go from drug design to safety testing.

Dr. Qingsong Zhu, Insilico Medicine’s chief operating officer told CAARE, “Insilico Medicine is contributing its efforts to reduce animal testing in drug development. We are using artificial intelligence technologies and in silico methods instead of animal testing to simulate and predict drug efficacy.”

Insilico_Logo_New.pngThe company is using deep learning to predict therapeutic properties of large numbers of drugs through the use of high-throughput experiments on human cells lines to supply gene expression data.

Deep learning is an outgrowth of computer science that uses data mining and pattern recognition to interpret large quantities of data. It’s the technology behind many things in use today, like powerful web searches, computer voice recognition, self-driving cars and deciphering the human genome.

The ability of artificial intelligence to analyze vast quantities of data while applying biological principles of gene expression allows it to determine the effects of compounds on cells, including therapeutic responses and possible side effects. 

Insilico Medicine is working to predict properties of drugs for a range of applications including treating common and rare diseases, aging, regenerative medicine and increasing response rates in cancer immunotherapy.

“There are just a few dozen drugs that are approved every year,” said Alex Zhavoronkov, Insilico’s CEO. “Basically one of the reasons why there is such a decline in research and development is because animal testing is not very representative of what the human outcome will be. Even though we are very similar to animals in many levels we are still very different, especially when it comes to complex systemic diseases.”

“We need something better, and something better is creating a virtual human to simulate the activity of many drugs on many tissues at once.” 


Citizens for Alternatives to Animal Research (CAARE), is a 501(c)(3) non-profit organization, established to highlight and promote research without animals.

Please donate to support CAARE’s mission to reduce animal suffering by disseminating information about the power and progress of research without animals. 

Click here to view this message online.