Would open-source, plug-n-play AI/ML tools be useful for the African drug discovery pipeline? The 263 applicants who applied for the AI/ML Tools in Drug Discovery Workshop hosted by H3D Foundation and Ersilia Open Source Initiative clearly think so.
The fortunate delegates who secured a spot in the 30-strong group from West Africa gathered at the Noguchi Memorial Institute for Medical Research at the University of Ghana from 9-12 September to receive hands-on training and develop new skills to accelerate their research projects.
The workshop aimed to teach the basic concepts of AI/ML and its challenges when applied to drug discovery; to create an awareness of the AI/ML tools that are available for scientific research, and to promote networking amongst different health-related disciplines.
At the intersection of science and art
Plug-n-play sounds easy, but AI/ML is not only a science, but also an art. AI/ML models, while powerful tools, are a bit like a needy pet: they need to be fed carefully-prepared data, trained to perform the required predictions, and monitored with considered attention.
Day 1 introduced concepts involved in working with chemical datasets before model training can begin, including: understanding AI/ML’s importance and usefulness in the drug discovery pipeline; connecting to sources of public data for screening/modelling; and exploring the chemical space, taking into consideration different molecular features.
Day 2 kicked off with an overview of the drug discovery pipeline by Guest Speaker Prof Richard Amewu. The focus of Day 2 was to understand the basics of how an AI/ML model is trained. While wet-lab scientists might not need to train their own models, they need to know what to look for when using the open-source tools. Our day ended with an inspiring tour of the University of Ghana Chemistry labs by our local hosts, showcasing the state-of-the-art facility and recent instrument acquisitions, and highlighting the potential for world-class drug discovery research in Ghana.
Day 3 began with an Introduction to Absorption, Distribution, Metabolism and Excretion (ADME), essential concepts for understanding the fate of a drug within a human body, by Guest Speaker Dr Arnold Forkuo Donkor. Thereafter, the facilitators demonstrated how to use generative models for chemistry, as well as AI/ML-aided searches of ultra-large chemical data sets. Participants explored expanding the chemical space around a molecule of interest, identifying related molecules and evaluating them from the perspective of ADME and antibiotic activity prediction.
Day 4’s keynote address was by Prof Samuel Kwofie on “Artificial Intelligence Assisted Drug Discovery: The Role of Big Data”. Thereafter, the key concepts of the workshop were recapped and applied to one of the participant’s current research project. This real-life application of the tools formed a bridge between the concepts explored in the workshop and the challenges faced in the labs.
Networking opportunities
With various scientific backgrounds amongst delegates (including chemistry, biology and computational sciences), forging bonds was a key part of the experience. Our programme included networking opportunities through fun activities such as the Canopy Walk in Legon Gardens, and salsa dancing at Afrikiko Leisure Centre. Our networking dinner was an opportunity to chat across tables of delicious local food, sharing professional goals and challenges.
We are grateful to our sponsor, Schmidt Sciences, and to our fantastic local hosts at University of Ghana and the Noguchi Memorial Institute for Medical Research. We’d like to give special thanks to our Guest Speakers and to our local coordinator Rawdat Baba-Adams.
Further workshops are planned for the future, so please stay tuned. In the meantime:
Check out our workshop materials here.
And keep in mind our free, online training opportunities:
H3D Foundation’s Massive Open Online Course “Introduction to Small Molecule Drug Discovery and Development”
Take a look at past editions of our H3D Foundation Webinar Series on YouTube. If you’d like to join us for future webinars, sign up here.
Also, connect with us:
H3D Foundation:
https://www.linkedin.com/company/h3d-foundation/
https://www.youtube.com/@H3DFoundation
Ersilia Open Source Initiative:
https://www.linkedin.com/company/ersiliaio
We hope to see you in an upcoming workshop soon!
Recent Comments