Novartis Taps Cloud HPC for Faster Drug Discovery & Better Science

One Drug - 10 years - $1 BillionOne Billion Dollars. $1,000,000,000. That’s a lot of zeros; nine to be exact. And that’s the estimated amount a pharmaceutical company invests to develop a new drug.

Perhaps even more dire is this process takes about 10 years, on average. In the broader view of human history, 10 years is nothing. But for someone who is sick, it may very well be a matter of life or death. It’s in this setting that we outline the perspective as to why accelerating research, discovery, and invention is so important.

Take Novartis Institutes for Biomedical Research, and the presentation they delivered at the 2014 AWS Summit in New York. In an important fight against cancer, the company leveraged the power and agility of Cloud computing to conduct 39 years of computing – in just 11 hours. They also created a computing system that would cost an estimated $44 million to build – all in the Cloud – for a cost of only $5,000.

AND most importantly: Novartis identified three promising compounds, which made their way into the next stages of drug design.

Steve Lister, Ph.D., Novartis Global Head of Scientific Computing

Steve Lister, Ph.D., Novartis Global Head of Scientific Computing

The Software Stack

Best of Show WinnerNovartis supports the work of a variety of researchers and scientist through high performance computing (HPC), and other technical computing capabilities. As part of a project to research a specific target protein associated with some cancers, the company turned to Cycle’s software to complete a massive computing workload in record time, and for an extremely low cost. Specifically, to support this class of workload, Novartis used the Bio-IT World Best of Show software CycleSever for workload placement, CycleCloud to orchestrate HPC environments across AWS, with the Chef 11 Server for continuous instance configuration, and HTCondor for job scheduling.


Cloud Software Stack

The Results

Three Compounds identified and synthesized for screening

First, this project for Virtual Screening for Cancer Research delivered valuable scientific results. As part of this run that lasted just 11 hours, doing 39 years of science, Novartis identified three compounds of interest. The scientific results are exciting as this data may ultimately help in designing a drug to fight cancer.

By leveraging Cycle Computing software and AWS Cloud infrastructure, Novartis was able to accomplish the same work faster, and for far less money.

  • $44 Million in infrastructure
  • 10 Million compounds screened
  • 39 drug design years in 11 hours for a cost of $4,232
  • 3 compounds identified for future work

This project was a highly successful project for Novartis. Not only did it deliver tangible results in three compounds, but it also proved out the business model, security model, and use case for scientific computing in the Cloud. Today Novartis boasts that their scientists have changed their way of approaching scientific computing problems. Instead of thinking about the types of questions they can ask with the computing power they have – they can now ask the un-restrained important questions, knowing Cycle software can help orchestrate the compute resources needed to get better answers, faster.

Share this: