New Way Now: Schrödinger recodes the rules of drug discovery with Google Cloud

*Featured in this video:* Shane Brauner, executive vice president and CIO of Schrödinger *Executive summary:* Schrödinger, a pioneer in physics-based computational modeling, migrated its platform to Google Cloud to transform the way new therapeutics and materials are discovered. Using Compute Engine to harness high performance computing resources like Cloud GPUs, Schrödinger now has access to near-infinite processing power on demand that previously required building and maintaining on-premises infrastructure. The ability to customize different resources also enhanced its utilization of services like BigQuery to bring together and analyze molecular properties for experimentation.The partnership has provided a quantum leap for Schrödinger’s scientific inquiry: Complex calculations that once took weeks now complete in hours, allowing experiments to run in a single weekend rather than months. This breakthrough speed and scale enables deeper, faster exploration of billions of possible compounds, with the potential to shorten multi-year discovery programs to months — significantly increasing the odds of identifying promising therapeutic candidates for patients. *Challenge:* Drug discovery projects require massive bursts of computing power that were difficult for Schrödinger’s existing on-premises infrastructure to supply, with peak needs that could surge up to 10,000 times greater than its standard capacity. Rather than building and maintaining more capacity that might not be necessary day-to-day, Schrödinger was looking for a way to expand its computational scale on demand while providing rapid, efficient, and accurate exploration. *Solution:* Partnering with Google Cloud, Schrödinger migrated its advanced computational platform to Google Cloud to increase its capacity and previous throughput. Harnessing high performance computing resources like Cloud GPUs, Schrödinger now has access to near-infinite processing power on demand without the burden or cost of building and maintaining on-premises infrastructure. The ability to customize resources also allows Schrödinger’s team to focus on delivering new ways to experiment, further enhancing its use of BigQuery and Compute Engine to bring together data and analyze molecular properties — all without having to invest in building and maintaining more capacity. *Results:* With Google Cloud, Schrödinger has delivered a quantum leap for drug discovery: Complex calculations now complete in hours instead of weeks, allowing researchers to explore billions of possible compounds in a single weekend — rather than months. This breakthrough speed and scale can dramatically shorten multi-year discovery programs to months by empowering researchers to rapidly explore billions of possible compounds and significantly increase the odds of finding promising therapeutic candidates. *Key takeaways and highlights from our interview with Shane Brauner, executive vice president and CIO of Schrödinger:* → “Google Cloud has been transformative for how we’ve been doing compute at Schrödinger. Good infrastructure is like a utility, and I love that about Google Cloud — it just works.” → “The delta between our typical run rate capacity on-prem versus what we could expand to at peak could be a factor of a thousand or 10,000. Our Schrödinger software scales beautifully horizontally and with Google Cloud, rather than a scientific inquiry taking months, we could do it in a weekend.” → “Before we transitioned our flagship product to GPUs, it took over three weeks to complete a calculation — that’s down to a matter of hours. Being able to run more experiments can really enable you to dig into another layer of understanding around the problem you're trying to solve. We've engaged with the Google product team and being able to explore and customize different machine types to leverage a flexible platform like BigQuery or Compute Engine, that enables us to experiment in new ways.” → “Being able to accelerate a drug discovery program from years to 10 months has a huge impact because time is not only money, but it's health for patients. We want to put a design-first approach to the drug discovery industry. It's a paradigm shift for the industry to do this discovery through simulation.” *Google Cloud products used:* Cloud GPUs, BigQuery, Compute Engine *Learn more:* → Schrödinger: Expediting drug discovery using the cloud https://cloud.google.com/customers/schrodinger → Schrödinger Expands Agreement with Google Cloud to Further Bolster Its Drug Discovery and Materials Science Initiatives https://ir.schrodinger.com/press-releases/news-details/2021/Schrdinger-Expands-Agreement-with-Google-Cloud-to-Further-Bolster-Its-Drug-Discovery-and-Materials-Science-Initiatives-02-25-2021/default.aspx