Eric Gilbert, PhD

• Worked with clients in pharmaceutical industry and government agencies to create ML and deep learning predictive models.
• Created transformer-based model pretrained on reaction SMILES and fine-tuned to predict reaction success.
• Created models to predict drug adverse events from structure using pre- and post-marketing data.
• Developed process to identify emerging trends in scientific literature.

Utilizing structure-based design and / or in silico predicted properties, designed compounds including those aimed at crossing the blood-brain barrier to reach a central nervous system target. Managed direct reports and external resources to execute on targets.

• Designed and prepared novel chemical matter across multiple chemical series’ of beta-secretase 1 inhibitors for treatment of Alzheimer's disease while Received Edison Patent Award as co-inventor of Verubecestat (MK-8931), Phase III clinical candidate for Alzheimer's disease.

• Championed use of predictive models to facilitate hypothesis-driven drug design while serving as team’s point of contact for chemoinformatic tools.
• Engaged in rotation within peptides group; focused on design and synthesis of macrocyclic peptide targets. Leveraged structure-activity relationships (SAR) and advanced peptide chemistry techniques to optimize pharmacokinetic properties and potency.

• Designed novel CB1 antagonist pharmacophore and prepared Giminabant, CB1 antagonist candidate for treatment of obesity.
• Co-inventor of Vicriviroc, CCR5 antagonist for treatment of HIV that progressed to phase II clinical trials. Received internal President's Award with team for efforts leading to discovery of Vicriviroc.