Regulome Summit 2025: A Milestone for Chromatin Biology and Drug Discovery
November 6, 2025
Regulome Summit 2025: A Milestone for Chromatin Biology and Drug Discovery
Seattle, WA - November 6, 2025 - The inaugural Regulome Summit was hosted by Talus Bio at the University of Washington on Monday, September 15, 2025. This one-day meeting brought together leading genome scientists, technologists, and AI experts to chart a collaborative path toward drugging challenging regulome targets.
The Summit opened with a session led by panelists from the Northwest Sarcoma Foundation, who spoke to the deeply rooted purpose of drug discovery: “every experiment you do, this matters.” Discussion flowed between the lived experiences of individuals with chordoma, the economics of funding rare disease research, and the growing role of non-profit foundations stepping in to advance those efforts. Panelists also addressed the unique challenges of the regulatory landscape and the importance of collaboration across industry, academia, non-profits, and patient organizations, such as working with the FDA to lower patient enrollment requirements for rare diseases, where a 300-patient clinical trial could represent nearly every known patient in the U.S.
It was deeply motivating to hear patients speak to the gap between the hypothetical “what’s possible now” in cutting-edge research versus “what might be available later” when it comes to turning proof-of-concept science into actionable treatments. Reflecting on this theme, Talus Bio CEO Alex Federation remarked that “when you sit with patients who’ve run out of options, it’s clear the issue isn’t that we don’t understand their disease, it’s that we haven’t been measuring the biology that actually drives it. That’s what regulome science makes possible.”
The AI/ML panel discussion was first kicked off by Dr. Jeff Leek, who outlined the major challenges of AI for biology, where the intersection of data and algorithms has been the main focus of most work, but lacks the crucial component of “interface” to elevate the data and algorithm from “interesting” to “insightful”. The key ingredients, he believes, are going beyond the hugely popular foundation model development to create actual products, and building trust, excitement, and engagement with those models.
Demonstrating real value from models is where the rubber will meet the road in the coming years, as the deluge of foundation models shakes out into what’s hype and what’s revolutionary. Panelists agreed that for AI/ML in biology, the value isn’t in the algorithms, but rather in the data used to train the algorithms and what the algorithm can produce. An interesting thread that came up was whether the value of AI/ML in biology was in foundation models at all, or whether the true value was in streamlining, improving, and automating operations through agentic AI.
A similar theme guided the next panel on epigenetics, opened by Prof. Steven Henikoff, who presented novel technologies to study the roles of DNA-binding proteins in cancer. His talk highlighted how proteins once thought to serve a structural function actually represent a fundamental vulnerability across cancer. The panel discussion that followed extended this theme, exploring how new layers of data, from single-cell technologies to whole-organism screens and regulome profiling, can power advances in AI models that reshape our understanding of cellular control.
“Our approach flips the typical AI paradigm,” said Dr. Lindsay Pino, CTO of Talus Bio. “Instead of using AI to guess biology, we generate the biological data that trains better AI. Regulome-scale proteomics gives us a window into fundamental biology that virtual cell models have been missing.”
The final discussion brought it all together to explore drugging the “undruggable”, combining AI/ML and epigenetics technologies to build therapeutics for diseases with previously untractable targets. Panelists agreed that “undruggable” is actually more accurately described as “unligandable”, where the challenge arises mainly from a lack of structure in the target protein. Typical tools used by drug hunters require structure, and exciting new AI/ML models like AlphaFold don’t work for most proteins in the regulome target classes.
The day closed with a happy hour for attendees to share their thoughts, connect with each other, and learn more. Posters demonstrating regulome profiling across a range of applications were displayed, including “Using proteomics to elucidate cancer killing mechanisms of anthracyclines” from Jenna Rosinski from Dr. Jay Sarthy’s lab (Seattle Children’s Research Institute), the winner of the Summit Poster Award.
About Talus Bioscience, Inc.
Founded in 2020 by Alex Federation, PhD, and Lindsay Pino, PhD, Talus Bio is pioneering the development of therapeutics that target transcription factors—the central regulators of gene expression long considered "undruggable." The Seattle-based company’s MARMOT platform combines AI, next-gen proteomics, and synthetic chemistry to profile, discover, and optimize small molecules that modulate transcription factor activity in live cells. With the world’s largest dataset of transcription factor activity and a team of experts in machine learning, biochemistry, and systems biology, Talus Bio is redefining what’s possible in drug discovery.
Media Contact:
Michelle Briscoe, COO & CFO
Talus Bioscience, Inc.