In the recent Nvidia GCT event, where the company announced some exciting developments to boost various industries, the pharmaceutical domain is not left behind. As Jensen Huang, the co-founder and CEO, noted in his keynote speech, understanding disease is one of the biggest challenges. Despite years of research and billions of dollars spent on drug discovery, 90% of the efforts fail. The COVID-19 situation has further thrown a curveball at the pharma industry.
Huang believes that using breakthroughs in computer science, one can begin to use simulations in in-silico methods to understand the biological machinery of the proteins that affect disease and search for lead candidate drugs before the slow process of in-vivo testing.
Sign up for your weekly dose of what's up in emerging technology.
There are many challenges that the field of drug discovery phases — firstly, it is hard to find a protein that is implicated in a disease; secondly, it is hard to find a molecule that can bind to protein to activate or deactivate it; and thirdly, getting the small molecule inside the cell is hard. To deal with these challenges, a plethora of spectacular technologies are being deployed from genomics to cryo-electron microscopic imaging and 3D reconstruction.
Another important method is to use data analytics which has been explored to screen billions of chemical compounds for fingerprints that might lead to a drug candidate. Also, deep learning and AI has been explored to generate and design new leads learning from known chemicals. Some of the other methods used are molecular dynamics, pathology, radiology, among others that have been used to develop or study the effectiveness of a drug. In fact, NLP has also been used to study publications and health records.
“Tackling the world’s most pressing challenges in healthcare requires massively powerful computing resources to harness the capabilities of AI,” said Jensen Huang, founder and CEO of NVIDIA, in his keynote.
Nvidia Clara Discovery For Making Life Saving Drugs
As Huang shared during his keynote speech, Nvidia Clara Discovery is a state-of-the-art suite of tools for scientists to discover life-saving drugs. With accelerated computing and AI, it promises to change the way biomedical research is conducted. It is a platform for imaging, genomics and patient monitoring which can be deployed anywhere — from embedded to the edge to every cloud. It aims to enable the healthcare industry to innovate and accelerate the journey to precision medicine. “Where no tools exist, we develop them,” said an elated Huang. He hopes that with this platform, researchers will be able to lead to the faster and more effective discovery of life-saving drugs.
Research in the healthcare domain demands extensive computing paradigms to carry research in areas such as personalised medicines and enhancing the quality of care. Many health institutions across the globe do not have the infrastructure to carry out the kind of research that is needed. Talking in his keynote, Huang noted that while the UK is the epicentre of healthcare research, it needs state-of-the-art computing infrastructure. Therefore, they built Cambridge-1, which will be the fastest supercomputer in the UK and top 30 in the world, with a speed of 400 PetaFLOPS.
“Cambridge -1 will host the UK AI and healthcare research collaborations with academia, industry and startups. Cambridge-1 will let them do experiments too large for their infrastructure,” he said. It will currently be adopted by companies and research institutions such as GSK, AstraZeneca, King’s College London and Oxford Nanopore Technologies and more.
Powered by 80 NVIDIA DGX A100™ systems, it will allow researchers and academics to tackle some of the most challenging problems in AI training, inference and data science workloads. While traditional supercomputers can take years to deploy, Huang mentioned that with modular DGX SuperPOD architecture, systems could be operational in a few weeks. NVIDIA is aiming to invest around $51.7 million in Cambridge-1, which is expected to be installed by the end of the year.
Apart from this, Huang also mentioned some of the installations that are already deployed across the globe.
- CDAC is going to be the most powerful supercomputer in India.
- Argonne National Lab is using DGX SuperPOD for computational drug screening.
- The University of Florida has a system built to train next-gen of AI researchers and students.
- Wallenberg AI and Software Research is making the largest AI infrastructure in Sweden.
Partnership With GSK
“Today we are also announcing a partnership of GSK and Nvidia to build the world’s first AI drug discovery lab,” said Huang in his keynote speech. GSK recently established a new AI Lab, based in London, which will leverage GSK’s genetic and genomic data to improve drug discovery and vaccine production.
Dr Kim Branson, GSK’s head of AI, said that GSK had accumulated more data this quarter than the entire history combined. It is the amount of data that no human can understand. “GSK believes that AI and computational biology when combined with our vast data sources, can break new ground in drug discovery. But we can’t do it alone. That is why GSK is partnering with Nvidia to build out the world’s first dedicated in-house drug discovery AI lab,” he added. They aim to explore AI methods and advanced computing platforms to unlock genetic and clinical data with increased precision and scale.
Apart from the announcements mentioned above, Nvidia is playing a key role in accelerating COVID-19 research. Using Nvidia’s scientific computing platform, organisations and institutions are being able to detect the disease better. It is also facilitating smart cameras to conduct temperature screenings and supercomputers that can scan one billion compounds in 12 hours in virtual drug screenings.
“Our end-to-end workflows are helping humanity’s greatest minds find breakthroughs throughout the entire COVID-19 spectrum – from detection and containment to mitigation and treatment, monitoring and tracking,” noted the blog post.
The developments mentioned above are indicative of the fact that Nvidia is taking the medical and healthcare domain seriously, and are enabling researchers with its state-of-the-art infrastructure and hardware to carry research at a scale never before seen.