Harnessing the power of AI for Alzheimer’s disease drug discovery at UAB System Pharmacology AI Research Center (SPARC)

SPARC Turns Alzheimer’s Research Into Action in 90 Days

When a Cell paper published on April 23, 2025, proposed a new transcriptional role for PHGDH in the amyloid pathology of Alzheimer’s disease, the UAB Systems Pharmacology AI Research Center (SPARC) didn’t treat it as the finish line—it treated it as a relay handoff.

Led by Sixue Zhang, Ph.D., associate director of SPARC, and Jake Y. Chen, Ph.D., SPARC’s director, the team mobilized its systems-pharmacology and generative-AI pipeline, powered by GeneTerrain Knowledge Maps (GTKM), to pressure-test the paper’s mechanism as a real-time case study.

In just 90 days, SPARC went from reading the paper to filing a U.S. provisional patent on its first batch of small-molecule candidates—demonstrating how a decentralized model can rapidly translate open science into actionable drug discovery.

From Mechanism to Molecules

Within those 90 days, SPARC’s team developed a suite of multi-scale, multi-modality deep learning AI models that not only identified novel PHGDH inhibitors but also established foundation models for future AI-driven drug discovery projects.

Instead of focusing on a single target, SPARC emphasizes process:

• Integrating the new mechanism into GTKM to understand its network context

• Generating CNS-suitable chemotypes

• Triaging candidates with multi-objective AI for potency, brain penetration, and early safety profiles

“Our aim wasn’t to claim a discovery; it was to show the relay works,” said Chen.“A strong finding appears in the literature, and our systems-pharmacology AI can turn it into make-ready molecules fast—then invite the community to help validate them.”

A Decentralized Drug Discovery Relay

As Chen explained in a recent Drug Discovery AI Talk,

“AI empowers startups, academic labs, and smaller organizations to drive therapeutic innovation. Agentic AI systems can automate experimental workflows.”

This decentralized relay flows from publication → design → bench:

1. Researchers publish a mechanism.

2. SPARC integrates and validates it, generating candidate compounds.

3. External partners—from academia, startups, or industry—conduct experiments to advance the best hits.

For the Alzheimer’s case study, SPARC’s provisional filing includes 18 synthetically accessible compounds prioritized for follow-up assays.

“I ran the end-to-end analysis—linking the literature to GTKM, generating and down-selecting compounds, and clearing early ADME/safety flags—so collaborators receive a short list worth making,”said Fuad Al Abir, a doctoral student in Chen’s lab who led the in-silico triage.

Next Steps

The continued design and development of these analogs will also draw on Zhang’s deep expertise in computer- and AI-aided drug discovery for Alzheimer’s disease. His recent publication in ACS Pharmacology & Translational Science describes a novel drug candidate demonstrating in vivo therapeutic efficacy for Alzheimer’s disease — further underscoring SPARC’s strength in AI-driven neurodegenerative research.👉 Read the publication here.

SPARC is now seeking partners to rapidly test the 18 compounds in:

• Enzymatic and cellular assays

• ADMET profiling

• Early in vivo models

A successful outcome could trigger company formation and accelerate IND-enabling studies toward clinical trials.

A Template for Open, Distributed Drug Discovery

SPARC envisions this as a template for decentralized drug discovery across chronic diseases:Treat each high-quality mechanism paper as a starting gun,run the AI-enabled relay to generate drug-like matter,and pass the baton to experimentalists for validation.

This model gives credit to the original authors for their mechanistic insights, positions SPARC as the systems-AI translator, and opens the door for rapid, collaborative validation across the biomedical community.