When venture capitalists showed little interest in backing two young founders pursuing an unconventional artificial intelligence-driven pharmaceutical approach, Xhuliano Brace decided to fund the venture himself.
Following a four-year stint working on artificial intelligence initiatives at Amazon, Brace departed the technology company to personally finance his ambitious vision. Drawing from personal savings and gains from digital trading activities, he poured a six-figure investment into Rhizome Research, a Seattle biotech venture focused on creating custom small drug-like molecules.
Established in the previous year, the company with five team members has recently stepped into the public eye. Beyond CEO Brace, who holds degrees in mathematics, computer science, and economics from the University at Albany, the startup’s leadership team features several accomplished professionals.
Co-founder and chief scientist Yiwen Wang brings a doctorate in chemistry from Carnegie Mellon University. Gregory Sinenka serves as chief technologist, contributing his physics doctorate and experience from both a European research facility and Johnson & Johnson. John Proudfoot, who previously held a director position in Boehringer Ingelheim’s U.S. Medicinal Chemistry Department, provides guidance as a scientific advisor.
**An Alternative Path for Pharmaceutical Development**
Instead of relying on conventional molecule-construction platforms, Rhizome developed its proprietary fine-tuned foundational model called r1. This technology operates as a “graph neural network” and underwent training using more than 800 million small drug-like molecular structures.
This methodology differs significantly from the widely-known RoseTTAFold model developed by the University of Washington’s Institute for Protein Design, which fundamentally relies on the amino acid components that form proteins.
The r1 model concentrates on the atomic components and chemical bonds that constitute a molecule along with its three-dimensional structure. This is where the graph concept becomes relevant — atoms function similarly to nodes in a graph while bonds resemble the connecting edges.
The company’s goal is to deliver fragment-based pharmaceutical development, generating small molecules specifically optimized to attach to client-designated targets. They guarantee that each drug candidate can be efficiently produced in laboratory settings and meets requirements for intellectual property protection.
Last week, Rhizome unveiled ADAMS, an open-source automated artificial intelligence platform that employs natural language commands for simulating interactions between biological molecules. The company also intends to release MolSim, a physics-driven simulation utilizing sophisticated free-energy calculations to forecast the binding strength between a small molecule and its target. However, MolSim will remain proprietary.
**Building a Seattle Innovation Center**
Rhizome has recently formed partnerships with wet laboratory facilities capable of confirming the practical effectiveness of its designed pharmaceutical candidates, while also exploring commercial partnerships.
Brace operates from Foundations, the Seattle-based entrepreneurial community created by entrepreneur and investor Aviel Ginzburg. While Rhizome’s remaining team members currently work remotely, the strategy involves relocating them to Washington state.
“I really want to make Seattle kind of a hub for small molecule drug discovery,” Brace said.
He referenced the Allen Institute, the Institute for Protein Design, and other Seattle-area institutions as significant contributors. The region also hosts numerous related pharmaceutical design ventures including Pauling.AI, Synthesize Bio, and Xaira Therapeutics, which maintains San Francisco headquarters with laboratory facilities in Seattle.
Brace expressed enthusiasm about pursuing a project with potential for substantial humanitarian impact and harbors no second thoughts about self-funding the endeavor. He maintains an optimistic outlook on artificial intelligence applications for molecular design, whether in healthcare or sectors like materials science and advanced
manufacturing.
“This is the most interesting problem space to be in,” Brace said.