What we're about

We are a growing community of greater New York City area scientists focused on the advances of in silico methods related to drug discovery, and how they increasingly impact and transform pharmaceutical research.

Our scientific interest covers a wide range of fields, including computational chemistry, medicinal chemistry, chemoinformatics, machine learning, molecular modeling, structural biology and protein design to cite a few.

NYAGIM is an inclusive forum for all academia and industry scientists interested in this particularly dynamic scientific interface, to connect, interact, learn and share.

We notably organize quarterly events where we gather for scientific seminars and discussions as well as for networking.

Upcoming events (1)

NYAGIM Event: Dr. Stephen MacKinnon (Cyclica)

Needs a location

NYAGIM is excited to welcome members back for our first in-person event since 2019!

Dr Stephen MacKinnon from Cyclica will present: "Proteome-Scale Drug-Target Interaction Predictions: Approaches and Applications"

The location is JLABS @ NYC (101 6th Ave 3rd floor) and the talk will be followed by a food and drink reception. Please note that we are strictly limited to 70 attendees and we are not able to broadcast this event online so it will be an in-person event only.

You can subscribe to the JJI/JLABS newsletter by following the link below:

https://jnjinnovation.com/subscribe

ABSTRACT

Drug Target Interaction (DTI) predictions have recently gained widespread popularity with advances in machine learning and publicly available bioassay datasets, such as BindingDB, DrugBank, PubChem and ChEMBL. Machine learning based strategies frame DTI predictions as a discriminative supervised learning problem, whereby combined pairs of features derived from the ligand (drug) and protein (target) are classified as a binding (positive) or non-binding pair (negative). Several literature-reported DTI models share this overarching classification strategy, but diverge in data representation, feature embedding strategies, training data quality thresholds, scale of underlying datasets, data balance, use of negative training examples, testing protocols and optimization targets. This presentation will review the overall structure of DTI prediction models, including strengths, limitations, and validation pitfalls based on our own first-hand experience developing Cyclica’s MatchMaker technology. Moreover, we will discuss the overall role of DTI models fit into today’s AI-led drug discovery workflows.

BIO

Stephen MacKinnon is a computational scientist and the Chief Platform Officer at Cyclica. He holds B.Sc. and Ph.D. degrees from the Universities of Waterloo and Toronto respectively, with research emphasis on computational biochemistry and structural biology. Stephen was Cyclica’s first staff scientist, where he led the design & implementation of their predictive technologies. Cyclica develops ML- and biophysics- based approaches to model a drug’s impacts on greater biological systems and applies these platforms to design and advance molecules that embrace the complexity of disease. In his current role, Stephen plans, coordinates and oversees the ongoing development of all in-house computational platforms to support drug discovery operations.

Past events (13)

NYAGIM Event: Dr. Pratyush Tiwary

This event has passed

Photos (25)