Generative AI transforms biology from science to engineering

Tuesday, August 1, 2023
AI
News
How do you envision healthcare driven by developments in technology and AI? In the same way pilots use simulators to practice for flights, surgeons can practice their craft in simulation with digital twins of organs. Doctors can create patient-specific models of an individual's organs to train before surgery. Medical students can practice on digital twins, too, with haptic feedback to reinforce training. Surgeons can also take advantage of guided surgery, with innovations like virtual tool tracking and telepresence. Generative AI is allowing researchers to train new AI models for chemistry and biology. This is unlocking a new understanding of molecules, proteins and DNA to uncover patterns in datasets, predict molecular reactions and optimize lead particles, representing a breakthrough for the pharmaceutical industry. Pharma companies will be able to build custom generative AI models through training and fine-tune with their vast proprietary data. NVIDIA is "powering healthcare solutions with accelerated computing." What does it mean in practice? Accelerated computing is core to the most advanced technologies including generative AI, simulation and real-time processing. In drug discovery, AI is being used to generate new proteins or molecules, predict structures and properties and virtually screen billions of therapeutic candidates, all in a computer vs the traditional experimental methods. This is reducing cost, speeding time to market and going beyond the human imagination. In genomics, accelerated genome analysis in population and cancer genomic studies can help identify rare diseases and bring tailored therapeutics to market faster, advancing the journey to precision medicine. For medical devices, AI can serve as an extra set of "eyes" for early detection and automatic measurement of anomalies, guiding surgeons and monitoring patients – making every hospital smart. Could you please give some examples of advances in drug discovery, genomics, medical devices and smart hospitals due to breakthrough technologies? NVIDIA BioNeMo is a cloud service for generative AI in biology, offering a variety of AI models for small molecules and proteins. With BioNeMo, pharmaceutical researchers and industry professionals can use generative AI to accelerate the identification and optimization of new drug candidates. Startup Evozyne used NVIDIA BioNeMo for AI protein identification to engineer new proteins with enhanced functionality that was experimentally validated. A joint paper describes the engineered proteins — one to potentially be used for treating disease and another designed for carbon consumption. Amgen enhanced its antibody design platform by customizing five proprietary generative AI models using BioNeMo, reducing development time from three months to just four weeks. Amgen then used the BioNeMo's AI models to predict the structure of the newly generated antibodies – all on NVIDIA DGX Cloud rather than lengthy and costly lab experimentation. NVIDIA Parabricks is a suite of AI-accelerated genomic analysis applications that enhances the speed and accuracy of the entire sequencing process, from gathering genetic data to analyzing and reporting it. A whole genome can be analyzed in 16 minutes vs. almost a day on CPU, meaning more people can be sequenced faster and cheaper so patients can get the help they need and researchers discover more about the thousands of diseases that are still untreated. Form Bio has recently integrated NVIDIA Parabricks into its computational life sciences platform, resulting in a 52% reduction in overall costs, enabling life sciences professionals to accelerate whole genome sequence analysis. Millions of medical devices are used every day across hospitals to enable robot-assisted surgery, radiation therapy, CT scans and more. NVIDIA Holoscan – a scalable, software-defined AI computing platform for processing real-time data at the edge – accelerates these devices to deliver the low-latency inference required for AI in a clinical setting. In a landmark step, doctors at Belgium-based surgical training center ORSI Academy brought NVIDIA Holoscan into the operating room to support real-world, robot-assisted surgery for the first time. Additionally, medical devices company Medtronic recently announced it will integrate NVIDIA healthcare and edge AI technologies into its GI Genius™ intelligent endoscopy module, developed and manufactured by Cosmo Pharmaceuticals. GI Genius is the first FDA-cleared, AI-assisted colonoscopy tool to help physicians detect polyps that can lead to colorectal cancer. Everybody talks about artificial intelligence. How will AI transform healthcare? In this new era of generative AI, AI is learning relationships, reasoning, predicting the structure and function of biomolecules and even designing novel therapeutics with desired properties. The results are groundbreaking – research shows that generative AI is transforming biology and chemistry from science to engineering, a breakthrough that is set to revolutionize the pharmaceutical industry. In other applications, AI in healthcare is delivering improvements across clinical ecosystems. With AI helping to streamline menial administrative healthcare tasks, clinicians are empowered to focus on giving the best possible care for their patients. Advancements in AI have reduced waiting times for radiology results from weeks to minutes. Breakthrough AI models are reducing the likelihood of patient readmission in hospitals, helping to avoid the downsides that come along with that, like worse outcomes and higher costs for both patients and hospitals. And what fascinates you most in potential use cases of AI in medicine and life sciences? The most fascinating use of AI in healthcare is the ability to unlock human understanding of biology. We've had the languages of biology and chemistry for decades, meaning we can read it. But now, with generative AI, we can begin to understand it, reason about it and generate new ideas to affect it. Generative AI is turning biology from science to engineering, which means we can shorten ten-year therapy development times, helping us personalize medicine, adapt to new diseases and notably reduce the cost, which increases access to modern medicine. Another fascinating and fast-moving area is surgery – minimally invasive, robotic surgery and telesurgery have enormous potential for human survival and reducing the cost of healthcare. Real-time AI is helping surgeons see, navigate, measure and eventually perform precision tasks, increasing ‌access to life-saving procedures, positive outcomes of complicated procedures and delivering procedures no human was capable of. The NVIDIA Inception program nurtures over 2000 healthcare startups developing tools to optimize operations, improve diagnostics, and elevate patient care. Can you give some of the examples you are most excited about? Healthcare startups are disrupting traditional models across the industry with innovative solutions that are greatly improving the quality and efficiency of care. Burgeoning AI startups are leveraging the technology to make sense of the wealth of medical data available to provide more accurate diagnoses, deliver optimized treatment recommendations and predict disease progression at a rate never before possible. For example, NVIDIA Inception member Insilico Medicine identified a drug candidate using its AI platform, and that candidate is now entering Phase 2 clinical trials to treat idiopathic pulmonary fibrosis, a relatively rare respiratory disease that causes a progressive decline in lung function. Another example is Inception member VinBrain, the creator of DrAid, which is the only AI software for automated X-ray diagnostics in Southeast Asia, and among the first AI platforms to be cleared by the FDA to detect features suggestive of collapsed lungs from chest X-rays. What should we do to avoid all these great healthcare/life sciences technologies being available only to those who can afford them? AI will help democratize healthcare by increasing access to care. The World Health Organization estimates that two-thirds of the world does not have access to diagnostic medical imaging. AI will improve access to radiology, improving detection rates and reducing barriers to care. As the pace of innovation increases, costs will go down, enabling the adoption of new systems into their workflows. Healthcare is a highly regulated industry making it much harder for technological breakthroughs to be implemented quickly in daily practice. Will technology manage to transform healthcare systems soon? AI is already transfering into big wins in healthcare – from chatbots to image quality checking to operating room monitors – none of which require regulatory approval. At the same time, the regulatory bodies are working hard to approve new technologies, and over 500 software-as-a-medical-device products are on the market today. It's a global imperative that a three-way partnership of healthcare, technology and regulatory bodies work together to educate, understand and implement the necessary safeguards to evolve and take advantage of AI's life-transforming tool. We've been talking about tech fixing ailing health systems for years. What gives you hope that this will really happen soon? Recent technological advancements have already demonstrated AI's positive impact in the clinical setting. Advancements in AI have reduced waiting times for radiology results from days to minutes. Breakthroughs in robotics have allowed patients to go home the same day from procedures that once required days of in-hospital recovery time.