Computational drug design, validated by experiment.
Axia Discovery designs novel peptide and small-molecule candidates with physics-based simulation and generative AI — and is built so every prediction is tested at the bench through CRO and academic partners.
Dr. Valentin Zhigulin
Founder & Chief Scientific Officer
Valentin is a computational scientist who has spent 20+ years building rigorous machine-learning and simulation systems for hard scientific problems — and was applying deep learning to science before it was mainstream.
An early deep-learning-for-science researcher. As a founding contributor to the Connectome project (2004) and among the first to apply convolutional neural networks to neuroscience imaging, his work on learning-based image restoration (ICCV 2007) has been cited 300+ times — deep-learning credibility that predates the current AI wave.
Trained at Caltech and MIT. Ph.D. in Physics & Computational Neuroscience (Caltech); postdoctoral research in computational neuroscience and machine learning (MIT). Earlier, physics at MIPT.
A decade-plus building large-scale, falsification-first quantitative systems. Leading quantitative research and machine-learning–driven strategies in institutional finance trained a discipline that now underpins Axia Discovery: large-scale model building, ruthless out-of-sample validation, and honest uncertainty quantification — the same rigor that separates a real computational edge from an overfit benchmark.
A builder. He left finance to design and self-fund two platforms end-to-end: nullary.ai, a negative-results intelligence engine that turns pharma's accumulated failure data into a queryable resource (120M+ negative findings, calibrated per-target models, full provenance), and Axia Discovery's in-silico design platform. Both are shipping products, not slideware — concrete evidence of the data-engineering and computational capability behind Axia Discovery's pipelines.
Method, in the open. Axia Discovery's computational methods are being written up for peer review (manuscripts in submission to Scientific Data and the Journal of Cheminformatics; preprints forthcoming), with the same evidence-first, uncertainty-reporting standard nullary.ai is built on.
Education
- • Ph.D., Physics & Computational Neuroscience — California Institute of Technology
- • Postdoctoral Fellow, Computational Neuroscience & Machine Learning — Massachusetts Institute of Technology
Leadership
Alesia Ivashkevich
Director, R&D (Australia)
Alesia Ivashkevich joined Axia as Director, R&D (Australia) with the intention of exploring the unique Australian ecosystem to enable effective translation of AI-assisted drug discovery into first-in-human clinical trials.
Her strong scientific background in radiation and cancer biology and extensive experience in end-to-end development of radiopharmaceuticals, combined with an inquisitive and strategic mindset, position her to shape Axia's innovative radioligand therapy (RLT) pipeline.
After completing her PhD at LMU Munich, Alesia joined a Sirtex-funded research program at the MRB laboratory at the Peter MacCallum Cancer Centre, focusing on the discovery of DNA-binding radioprotectors. Her time with the Radiation Oncology department at Canberra Hospital provided valuable research and administration experience within a clinical environment.
Most recently, as Director, TME Research at Telix Pharmaceuticals, Alesia gained extensive experience in the end-to-end development of radiopharmaceuticals — from building the strategy for a new pipeline to advancing assets into preclinical and clinical development, guided by strong scientific rationale and alignment with TGA/FDA requirements. She demonstrated leadership through her support of the ImaginAb acquisition and the Regeneron deal.
Alesia excels at building productive relationships with scientific, commercial and clinical leadership, key opinion leaders, collaborators, and strategic partners, staying at the forefront of cancer targets and novel developments in the RLT space.
Alex Harwig
Head of Translational Biology & Platform Development
Alex is a translational biotechnology leader with nearly three decades of experience turning early biology into executable drug-development programs across oncology, molecular imaging, gene therapy, protein engineering, radiotherapy, and AI-enabled discovery.
A discovery-to-clinic operator. He has helped early-stage companies build molecular biology, therapeutics, assay-development, imaging, and translational capabilities, contributing to programs that moved from research concepts into IND-enabling work and clinical execution. His work spans target validation, assay cascade design, biodistribution, DMPK/ADME, nonclinical strategy, CRO management, and FDA-facing development.
A platform builder. Alex has built and led scientific platforms at the interface of molecular biology, translational pharmacology, and computational decision-making, including synthetic biomarker systems, receptor-targeted payload strategies, and AI-native scientific/IP intelligence through FYLED.
A scientist-inventor. His work includes 25 publications, 1,200+ citations, and co-inventorship on 30 patent families, including technologies in inducible gene-expression systems, synthetic cancer-specific promoters, molecular imaging, and therapeutic payload localization.
At Axia Discovery, Alex leads translational biology and platform development, connecting computational peptide design to the assay, conjugation, radiochemistry, in vivo, CRO, and early clinical strategy required to build radioligand therapy programs.
Education
- • Ph.D., University of Amsterdam
- • B.Sc., Biology and Medical Laboratory Techniques — Saxion University
- • Research training — University of Wisconsin–Madison
A de novo design platform for selective peptide therapeutics.
Axia designs candidate molecules computationally — physics-based simulation plus generative AI — against receptor targets that are clinically validated but hard to drug selectively, then confirms them at the bench. Candidates are engineered for selectivity against the closest receptor relatives, and advanced only as the data supports.
Pipeline — preclinical (in-silico-designed → wet-lab validation underway)
Ten patent-pending programs across four therapeutic areas, led by oncology radioligand therapy.
Oncology · radioligand therapy — the lead franchise
De novo cyclic-peptide targeting ligands for tumor-selective receptor antigens beyond today’s approved radioligands — AX-RLT-001, AX-RLT-002, and AX-RLT-003 — built for ¹⁷⁷Lu / ²¹²Pb / ²²⁵Ac therapy with a ⁶⁸Ga companion diagnostic.
CNS / neuropsychiatry
Selective peptide modulators of validated CNS GPCRs (AX-CP-001, AX-CP-002) — addressing mood and rare epilepsy/obesity via mechanistically distinct agonist and antagonist designs from one chemistry.
Cardiometabolic & renal
Designed cyclic-peptide programs — AX-CP-003 (heart failure / PAH), AX-CP-004 (obesity / metabolic), and AX-CP-005 (secondary hyperparathyroidism in CKD-MBD).
Every program is at computational-design or early CRO-validation stage. We report what the models predict and what the bench has confirmed — we do not claim pipeline-stage milestones the data does not yet support.
How we work
Computation generates and prioritizes; experiment decides. Candidates are designed in silico — physics-based simulation plus generative AI — then confirmed for binding, selectivity, and developability through CRO and academic partners before a program advances. We report what the models can and cannot do.
Interested in partnering, licensing, or advancing a difficult target?
Get in Touch