Website Danish Technical University (DTU) and Novo Nordisk
Job description
Are you passionate about understanding the interplay between AI-based protein design, automation and cutting-edge data science? Would you like to be part of a new research collaboration between DTU and Novo Nordisk? Then you could be our new PhD student. Read on to find out more! About the PhD program – AI-driven closed-loop design of protein binders The Danish University of Technology (DTU) and Novo Nordisk have entered into a strategic partnership and created an interdisciplinary joint research program aimed at solving one of the biggest challenges faced by the life sciences industry: closed-loop design and optimization of biologics. The research program will build on recent advances in protein design, automation and multi-parameter optimization to create a closed-loop pipeline capable of rapidly designing binders for any target and optimizing for exploitability. The program is rooted in DALSA (DTU Life Sciences Automation Arena), a new interdisciplinary center at DTU dedicated to bringing together life sciences, automation and data-driven innovation to revolutionize research and development. The program combines expertise from five DTU departments (Structures, Bioengineering, Computing, Electronics and Health Technologies) with mentors from DTU and Novo Nordisk. What you can expect As a PhD student, you will be immersed in a dynamic and collaborative research environment that includes PhD, postdoctoral and master’s students working on research-related projects. Expert mentors from DTU and Novo Nordisk. The lead mentor is affiliated with the Bioinformatics department at DTU Health Tech and the associate mentor is based at Novo Nordisk. This team of mentors ensures the highest level of academic expertise to support projects and programs. There are opportunities to attend conferences, symposia and networking events to share and enhance your research. You will play a key role in driving innovation and contributing to the transformation of biologics development. Research Program Recent advances in generative artificial intelligence (genAI) have revolutionized the design of molecular binders, especially within the framework of ab initio and closed-loop optimization. By integrating AI-driven approaches in a closed-loop system, we can iteratively improve adhesive design to achieve high affinity and specificity. Key components include model generation: generating an initial set of diverse protein backbones using advanced models such as diffusion models. Sequence refinement: explore sequence space using generative tools such as ProteinMPNN to identify sequences with improved properties. Validation: Using AlphaFold2 and other so-called oracles, candidate sequences are further enriched to identify proteins with appropriate folding, target interactions and biophysical properties. Optimization: using “design of experiments” and Bayesian optimization techniques to effectively guide sequence selection and strike a balance between exploration and exploitation. The experimental results of the combined assays are fed back into the system to enhance the design process in subsequent iterations. This integrated approach aims to significantly improve the efficiency and success rate of developing high-affinity molecular binders for industrial applications. Purpose The purpose of this PhD project is to optimize the adhesive design process through closed-loop optimization with an emphasis on efficiently achieving high affinity and specificity binders. This project will address the following research questions: create “0-shot oracles” (artificial intelligence models) that can be used to screen sequences and structures from generated models to improve hit rates. Develop bias and/or training strategies to tune oracles based on existing experimental data. Explore and implement model uncertainty prediction strategies in algorithms. Integrate these methods in a closed-loop optimization paradigm.
offer requirements
We are looking for a highly motivated candidate with a master’s degree in bioinformatics, computer science, or a similar field. In particular, we are looking for candidates with expertise in deep learning programming for generative models, language models, and/or inverse folding models. A solid background in protein sequence, structure and function is also required. You must have a two-year master’s degree (120 ECTS points) or a similar degree at an academic level equivalent to a two-year master’s degree.
offer benefits
DTU is a leading technical university recognized globally for excellence in research, education, innovation and scientific advice. We offer you a rewarding and challenging job in an international environment. We pursue academic excellence in an environment of respect for colleagues, academic freedom and responsibility. Remuneration and terms of appointment Appointments will be based on a collective agreement with the Danish Federation of Professional Associations. Stipends will be agreed with the relevant trade unions. The term of appointment is 3 years. Click here to find out more about careers at DTU. For more information please contact Carolina Barra Quaglia ([email protected]). For more information about DTU Health Tech, please visit: www.healthtech.dtu.dk/. If you are applying from abroad, you can find useful information about working in Denmark and at DTU in DTU – Moving to Denmark. In addition, you can attend our free monthly seminar “PhD Relocation to Denmark and Launching ‘Zoom’ Seminar” to learn about all the practical aspects of relocating to Denmark and working as a PhD at DTU.
To apply for this job please visit dtu.dk.