AI to leverage data and close the loop between patient data and R&D
Posted 15th October 2020 by Nick Noakes
Ahead of the Global Pharma R&D AI, Data Science and Informatics Summit we sat down with Claire Biot to talk about her work to provide life science companies & professionals with a scientific and business platform to imagine sustainable innovations, capable of improving patient & physician experience in the age of precision medicine.
From your perspective, how is AI revolutionizing pharmaceutical R&D and drug discovery?
In many ways, AI is a natural evolution of things we’ve always done in life sciences. We have collected massive amounts of data and applied statistical models to help us understand the data. We have furthered our understanding using predictive techniques, machine learning and simulation to see what we couldn’t see before. Now, using these datasets and models as a foundation to predict efficacy and safety, we can train AI algorithms to explore possibilities we may not have considered.
For example, with generative techniques, AI algorithms can be initiated by scientists with ideas of potential molecules to target disease and generate new molecules without human bias and predict how well they perform. Then repeat this over many generations, honing in on those that meet desired criteria. Of course, much like driving directions, you want a human in the loop, using data and algorithms to navigate you around traffic bottlenecks in ways you might not have known. For the scientist, this isn’t so much artificial intelligence as augmented intelligence helping to navigate the challenges of efficacy, safety, synthesizability and sustainability.
How can Dassault Systèmes’ BIOVIA software portfolio support scientists and modellers in their work?
To fight a disease, you need to understand it to discover how to tackle it and identify the best target. Once you have identified the target, you need to find the best candidate and ensure it works as expected.
Pharmaceutical and Biotechnology companies are challenged with managing open scientific innovation and accelerating the time of therapeutic design and development, while driving down costs. This requires connecting the dots between people, resources, processes, and data. Dassault Systèmes’ 3DEXPERIENCE platform enables scientists to work in partnership and efficiently share scientific data, accelerating decision support to move the best candidates forward.
The platform provides a unique way to unify predictive science (modelling & simulation, assisted by AI as described in my previous answer) and experimental execution at the bench: analyzing the results that are generated at the bench and comparing them to models allows researchers to rapidly optimize therapeutic efficacy and safety. It also provides an early prediction of potential downstream manufacturability issues, to guarantee rapid formulation of efficacious, stable therapeutics and speed up time to market.
What developments in the platform are you most proud of?
Joining forces with Medidata last year allows us to create the first end-to-end scientific and business platform, from research & discovery up to commercialization, through preclinical development, clinical testing and manufacturing. Our belief at Dassault Systèmes is that virtual worlds extend and improve the real world. Medidata’s clinical & real-world data & advanced analytics will strongly combine with the power of modelling and simulation to catalyze the next generation of patient-inclusive therapeutics: by leveraging data, we indeed close the loop between patient data and research and discovery.
What are the current major challenges in pharma data science and informatics, and what is Dassault Systèmes’ role in solving them?
Data science without models is of no use. But modelling biology is complex. We need to rely on reality and real-world observations to create models that mimic reality. These models, that we can simulate, we call them virtual twins – we can create virtual twins of a molecule, a cell, an organ, and imagine game-changing innovations at the age of precision medicine. But to calibrate these twins, we need to collect real-world data.
In what ways do you hope to be able to assist with therapeutic discoveries and personalized medicine in the future?
From its DNA to its organs, the human body holds complex mysteries yet to be uncovered by science. Inefficiencies in today’s world of research and medicine, compounded with the need for more precise, affordable patient care, have made the healthcare industry ripe for technological innovation that can transform how therapies are discovered, developed, commercialized, produced and used.
Imagine being able to understand, model, search, test, and treat a human body as precisely, safely and effectively as we already can today for a plane, a car or a building. We can transform how people are cured and help them live a better life.
Claire Biot is VP Life Sciences Industry at Dassault Systèmes. Ton Van Daelen Portfolio Lead, Scientific Informatics, Dassault Systèmes BIOVIA will be speaking about ‘Combining Machine Learning and Molecular Modeling for Enhanced Active Learning in Small-Molecule Drug Design’ at the Global Pharma R&D AI, Data Science and Informatics Summit.
The 4th Global Pharma R&D AI, Data Science and Informatics Summit will focus on the advancements in AI, informatics and data science for drug discovery & clinical development.
Click here to see what else is on the agenda.
Leave a Reply