AI-driven PIC™ Design
We design novel proximity-inducing compounds (PICs™), to treat diseases with high unmet medical need. Our patent-filed approach uses active machine learning to predict biomolecular interactions and generate new chemical entities.
The wholly-owned pipeline is focused on degrader development in oncology and CNS disorders, while we are open to partnerships across all indication areas.
Watch the video to learn more.
Machine Learning
We are exploiting cutting-edge machine learning to design PICs™. Traditional computational approaches have significant limitations with regards to speed and accuracy. Machine learning has demonstrated the ability to outperform these methods in the fields of natural language processing, machine vision and protein folding. Engineering new medicines requires optimizing multiple parameters, including activity, solubility, permeability, synthesis, toxicity, stability, and many, many others.
As machine learning requires high-quality, domain-specific data, our solution utilizes active learning to identify key data needs and generates them in-house. This closed loop comprising the entire design, make, test process together with proprietary databases and highly curated public datasets enables our system to sustain industry leading performance. The more iterations completed, the better the results.
Proximity-Inducing Compounds
Proximity-inducing compounds (PICs™) give scientists a unique opportunity to control biological function. Modifying local protein concentrations unlocks novel mechanisms of actions targeting previously untreated diseases.
What are PICs™?
Proximity-inducing compounds are a novel class of therapeutic modalities able to hijack naturally ocurring biological machines. All cells have the ability to express, degrade and deliver proteins to various compartments. Processes such as autophagy, folding via chaperones, lysosomal and ubiqutin-mediated proteolysis all affect homeostasis in protein levels. PICs allow us to modulate the activity of these processes and hence provide unprecedented potential to further expand therapeutic horizons.
Celeris Therapeutics is the pioneer in machine learning-driven PIC™ design focused on targeted protein degradation.
Targeted Protein Degradation
Mono- and bivalent degraders initiate the process of targeted protein degradation. They are shifting paradigms from a focus on inhibition towards removal of pathogenic proteins. These molecules induce protein-protein interactions, and overcome the challenges traditional small molecules were facing throughout the last century.
Degrader technologies are clinically validated and demonstrate superior selectivity, activity and therapeutic window.
Our platform Celeris® One exploits the potential of PICs™ with accurate predictions of biomolecular interactions and the expansion of chemical space.
Active machine learning serves as an enabler in this new area.
One Platform
Enables Rational PIC™ Design
Celeris One™
Celeris One™
PIC™ Design
Xanthos is the platform’s design engine. Starting with a protein target sequence, we determine 3D structures, predict ligand binding, generate linkers, predict ternary complexes and refine PIC™ selection through several layers of filters including molecular dynamics and QSAR.
From Data to Hit
Automated Synthesis Lab
Hephaistos is the platform’s synthesis engine. Machine learning-driven retrosynthesis and a robotized synthesis lab create real-world, custom chemical matter for testing in our automated biology lab (Hades). Our facility includes all necessary organic synthesis equipment, such as fume hoods, evaporators, HPLC, LC-MS and microfluidic synthesis.
From Hit to Compound
Biological Validation
Hades is the platform’s biological information engine. It generates experimental data obtained from our automated lab which we began constructing mid 2021. We exploit several, different biochemical assays and phenotypic screening protocols, and have in-house capabilities in cell culture, high-content imaging, functional assays and molecular biology.
From Compound to Lead
The PIC™ Engineering Company
We are CelerisTx
We are an international team of highly skilled experts from disciplines as computational chemistry, medicinal chemistry, biology and machine learning.
Our team is complemented by highly esteemed advisors and investors at the intersection of artificial intelligence and life sciences.
One Team. One Vision.
LEADERSHIP
Christopher Trummer
CEO, President and Co-Founder
Jakob Hohenberger
CFO/COO, Director and Co-Founder
Dr. Markus Müllner
CSO
Noah Weber
CTO
BOARD OF DIRECTORS
Marius Swart
Director
Karel Kubias
Director
Dr. Peter Ho
Director
Petr Šrámek
Observer and Advisor
Dr. Ronjon Nag
Observer and Advisor
Scientific Advisors
Prof. Dr. Jean-Louis Reymond
Former Dir. of Chemistry and Biochemistry at the University of Bern.
Assoc.-Prof. Dr. Johannes Kirchmair
Head of Comp. Drug Discovery and Design at the University of Vienna.
Prof. Dr. Ola Spjuth
Professor at the University of Uppsala.
Prof. Dr. David Selwood
Professor at the University College London.
Prof. Dr. Stefan Knapp
Professor at the Goethe University Frankfurt.
Open positions
Our Investors
Our Supporters
