Postdoc in AI-driven design of environment-responsive antibodies and nanobodies

Are you interested in using machine learning to design proteins that respond to their environment and control biological function with high precision?

We are looking for a postdoc for a 4-year position starting 1st of September 2026 (or soon thereafter) in the Antibody Technologies Group at DTU Bioengineering, led by Professor & Head of Section Andreas Hougaard Laustsen-Kiel.

In this project, you will work at the interface of machine learning and protein engineering, computationally designing antibodies, nanobodies, and binding proteins with condition-dependent activity. You will develop deep learning and generative models for binder design, work closely with wet lab scientists who express and test your designs, and collaborate with leading international partners. The position offers the opportunity to take ownership of your project and contribute to new strategies for controlling protein function in complex biological environments.

The research project: PROTÆCT

This postdoc position is part of the Villum Foundation funded project, PROTÆCT: Precision Regulation Of Toxin Activity through Environmentally-Controlled Targeting, which aims to uncover how protein-based toxins can be controlled using binding proteins that respond to environmental cues such as pH and metal ions.

Most engineered proteins function under static conditions, whereas biological systems are dynamic. PROTÆCT addresses this by developing binding proteins with condition-dependent activity, enabling selective modulation of protein function in defined microenvironments.

You will computationally design antibodies and nanobodies whose binding changes with environmental conditions. This could e.g. enable the development of proteins that are only active in specific tissue environments or are activated when conditions change during a fermentation process, with potential applications in targeted therapeutics, biosensing, and synthetic biology.

The project integrates computational design with advanced in vitro selection technologies. You will play a central role in developing the computational models and in linking these to how environmental factors influence protein–protein interactions.

The research group

You will join the Antibody Technologies Group at DTU Bioengineering, working at the interface of toxinology, antibody engineering, and protein science. The group has a strong track record in developing recombinant antivenoms, environment-responsive antibodies, and translational biotechnological solutions.

We are an international and interdisciplinary team combining experimental and computational expertise. You will be expected to take ownership of your project and work independently, while contributing to a collaborative environment where ideas and expertise are shared across the team.

The group offers a setting with strong academic and industrial connections, providing opportunities to engage with leading collaborators and contribute to research with both fundamental and translational impact.

Responsibilities and qualifications

The postdoc project will focus on in silico design, discovery, and characterization of environment-sensing binding proteins targeting natural toxins.

Key tasks may include:

• Designing and implementing state-of-the-art deep learning architectures for generating binding proteins that sense their biochemical environment
• Developing and optimizing generative models for de novo design of binding proteins that sense their biochemical environment
• Collaborating closely with wet lab scientists who will express and test the designs that have been created
• Publishing results in impactful journals and contributing to future grant applications
• Scientific project management and co-leadership
• Mentoring and training PhD students and MSc/BSc students

We are looking for a candidate who:

• Holds a PhD in Computer Science, Machine Learning, Artificial Intelligence, Computational Biology, or a closely related field
• Has strong theoretical and practical experience in deep learning
• Has hands‑on experience developing generative models
• Is highly proficient in Python, PyTorch and/or JAX
• Has experience training large-scale neural networks on HPC or GPU clusters
• Has experience with representation learning and sequence or structural modeling
• Has a strong publication record in relevant AI/ML venues or interdisciplinary journals
• Demonstrates the ability to work independently and drive technical innovation
• Demonstrates the ability to manage projects and mentor students

It will be considered an advantage if you have:

• Experience with protein language models (e.g., ESM, ProtT5) and/or antibody specific language models (e.g., AntiBERTy, AbLang2)
• Experience with structure prediction fram