The Tubiana Lab leverage advanced machine learning and statistical physics to decode complex biological structures and engineer functional proteins. By developing generative AI models and data-driven computational algorithms, the lab accelerates biotherapeutic discovery, enabling the rational design of custom proteins and molecular interactions with high specificity.

Research Focus

• Computational Structural Biology: Applying statistical physics and machine learning to model protein structures, dynamics, and molecular interactions at atomic resolution.
• AI-Driven Protein Design: Developing generative artificial intelligence models to design novel functional proteins, enzymes, and therapeutic scaffolds from scratch.
• Machine Learning for Biomolecules: Building specialized deep learning and statistical inference frameworks to analyze large-scale biological datasets and predict molecular function.

Capabilities

• Generative AI & Machine Learning Modeling: Custom machine learning architectures tailored for biomolecular sequence, structure, and interaction predictions.
• Statistical Physics & Computational Frameworks: Physical-statistical modeling methods to evaluate protein stability, folding landscapes, and conformational changes.
• In-Silico High-Throughput Screening: Computational pipelines for rapid screening and optimization of candidate protein leads prior to experimental synthesis.

Partnership & Opportunities in Different Industries

• Biopharma & Therapeutic Protein Discovery: Partnering with biotechnology and pharmaceutical companies to design novel antibody candidates, peptide drugs, and targeted protein therapeutics.
• Industrial & Synthetic Biology: Collaborating with agtech and industrial biotech partners to engineer custom enzymes and biocatalysts with enhanced stability and activity.
• Software & Computational Bio Tools: Licensing proprietary machine learning algorithms and computational tools to life science software and platform companies.
• Potential Market Segments: ICT & Media (Machine learning & AI), Therapeutics (Drug Discovery, Bioconvergence), Research Tools (Research Software), Synthetic Biology.

Selected Publications

• Generative models and statistical physics approaches for protein sequence and structure modeling – Demonstrating machine learning frameworks for predicting structural conformations and functional mutations.
• Machine learning architectures for de novo protein design – Outlining data-driven computational strategies to design novel, stable protein geometries for biomedical applications.

Contact:

Email: jeromet@tauex.tau.ac.il