Accelerate Discoveries in the Field of Medical Research
Current technologies are already making it possible to accelerate discoveries: few examples.
Current technologies are already making it possible to accelerate discoveries, particularly in the field of medical research. This acceleration will certainly be accentuated with technologies that are developing at all speeds, such as decision-making algorithms derived from artificial intelligence, the ability to better identify genes, which will make drugs more personalized and better suited. If we look further ahead, future technologies like quantum computing could make major discoveries. We can very well imagine that in a few years, there will be as many treatment protocols as there are sick people, but also more anticipation and prevention.
High Performance Computing
Today, if we take the example of the use of supercomputers or the use of AI to analyze data, technology is an indispensable tool in medical research. We can cite one example with the COVID-19 High Performance Computing Consortium (43 members and 98 projects). It is a public-private effort initiated by several institutions of which IBM. It brings together the US Federal government, industry, and academic leaders to provide access to the world’s most powerful high-performance computing resources in support of COVID-19 research. The scientists are working on projects related to COVID-19 to help them identify potential short-term therapies for patients with the virus.
Since its launch in March 2020, the Consortium’s compute capacity has nearly doubled to 600 petaflops (million billion floating operations per second), up from 330 petaflops in March. Scientists have access to more than 50K GPUs, 6.8m cores and 165K nodes.
Together, the Consortium has helped support more than 95 research projects, such as:
- Understand how long respiratory droplets persist in the air: This research by a team at Utah State University simulated the dynamics of indoor aerosols, providing insight into how long respiratory droplets linger in the air. They found that droplets from breathing linger in the air much longer than previously thought, due to their small size compared to droplets from coughs and sneezes.
- Research on the reuse of drugs for potential treatments: A Project by a Michigan State University team looked at data from about 1,600 FDA-approved drugs to see if there are possible combinations that could help treat COVID-19. They have found promising combinations.
- Examining the Potential of Indian Herbal Medicines: Research by Novel Techsciences in India screened natural compounds derived from plants of 55 Indian herbal remedies to identify compounds with antiviral properties that could be used against eight SARS-CoV proteins -2.
Artificial Intelligence
Artificial intelligence is also critically important to accelerate research. We can cite two technologies: Corpus Conversion Service and Corpus Processing Service. To help researchers quickly access structured and unstructured data, IBM Research has developed a cloud-based AI research service that has enabled a corpus of thousands of articles to be ingested. (158,524 as of January 18, 2021) from the COVID-19 Open Research Dataset (CORD-19) and databases licensed from DrugBank, Clinicaltrials.gov, and GenBank. This tool uses advanced AI, allowing users to make specific queries across the many articles and extract essential knowledge about COVID-19 — including embedded text, tables and figures. This service helps innovation in various sectors such as materials science, insurance and drug discovery. We can easily imagine that it is impossible for a scientist to read 150 000 papers in a few minutes. Even if you only take 5 minutes to read an article that would be 12 500 hours. We only have 8,760 hours in a year. Both technologies are already widely used in materials science, automotive and energy industries. Corpus Conversion Service can ingest 100 000 PDF pages per day (even scanned documents) on a single server — then train and apply advanced machine learning models that extract content from those documents with great precision. Then, we can ask different queries such as which drugs have been used so far and what are the outcomes ? Identify new, reported risk factors ?
What’s next ?
We will also be able to address the future of medical research with quantum computing, which should eventually allow the synthesis of new therapeutic molecules, thus making it possible to discover new drugs, faster and more precise medical diagnostics.
Indeed, if we take the example of the caffeine molecule, it is a relatively simple molecule with only 24 atoms. And yet, even if you took the most powerful computers on the planet you would be unable to solve quantum mechanical equations to predict its structure and reactivity without making approximations. This will require more than 1048 binary units. A quantum computer which obeys the same laws as this molecule, could allow simulations much more faithful to reality. Quantum computing will improve the performance of artificial intelligence allowing the healthcare world to reach a level of personalization unattainable today.
As Nobel Laureate Richard Feynman said, “Nature is not classical, damn it, and if you want to simulate nature, you better make it with quantum mechanics.”
It’s a new way to do research that allows accelerating discoveries.
