A Variational Perspective on Generative Protein Fitness Optimization

Lea Bogensperger, Dominik Narnhofer, Ahmed Allam, Konrad Schindler, Michael Krauthammer
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:4700-4712, 2025.

Abstract

The goal of protein fitness optimization is to discover new protein variants with enhanced fitness for a given use. The vast search space and the sparsely populated fitness landscape, along with the discrete nature of protein sequences, pose significant challenges when trying to determine the gradient towards configurations with higher fitness. We introduce Variational Latent Generative Protein Optimization (VLGPO), a variational perspective on fitness optimization. Our method embeds protein sequences in a continuous latent space to enable efficient sampling from the fitness distribution and combines a (learned) flow matching prior over sequence mutations with a fitness predictor to guide optimization towards sequences with high fitness. VLGPO achieves state-of-the-art results on two different protein benchmarks of varying complexity. Moreover, the variational design with explicit prior and likelihood functions offers a flexible plug-and-play framework that can be easily customized to suit various protein design tasks.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-bogensperger25a, title = {A Variational Perspective on Generative Protein Fitness Optimization}, author = {Bogensperger, Lea and Narnhofer, Dominik and Allam, Ahmed and Schindler, Konrad and Krauthammer, Michael}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {4700--4712}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/bogensperger25a/bogensperger25a.pdf}, url = {https://proceedings.mlr.press/v267/bogensperger25a.html}, abstract = {The goal of protein fitness optimization is to discover new protein variants with enhanced fitness for a given use. The vast search space and the sparsely populated fitness landscape, along with the discrete nature of protein sequences, pose significant challenges when trying to determine the gradient towards configurations with higher fitness. We introduce Variational Latent Generative Protein Optimization (VLGPO), a variational perspective on fitness optimization. Our method embeds protein sequences in a continuous latent space to enable efficient sampling from the fitness distribution and combines a (learned) flow matching prior over sequence mutations with a fitness predictor to guide optimization towards sequences with high fitness. VLGPO achieves state-of-the-art results on two different protein benchmarks of varying complexity. Moreover, the variational design with explicit prior and likelihood functions offers a flexible plug-and-play framework that can be easily customized to suit various protein design tasks.} }
Endnote
%0 Conference Paper %T A Variational Perspective on Generative Protein Fitness Optimization %A Lea Bogensperger %A Dominik Narnhofer %A Ahmed Allam %A Konrad Schindler %A Michael Krauthammer %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-bogensperger25a %I PMLR %P 4700--4712 %U https://proceedings.mlr.press/v267/bogensperger25a.html %V 267 %X The goal of protein fitness optimization is to discover new protein variants with enhanced fitness for a given use. The vast search space and the sparsely populated fitness landscape, along with the discrete nature of protein sequences, pose significant challenges when trying to determine the gradient towards configurations with higher fitness. We introduce Variational Latent Generative Protein Optimization (VLGPO), a variational perspective on fitness optimization. Our method embeds protein sequences in a continuous latent space to enable efficient sampling from the fitness distribution and combines a (learned) flow matching prior over sequence mutations with a fitness predictor to guide optimization towards sequences with high fitness. VLGPO achieves state-of-the-art results on two different protein benchmarks of varying complexity. Moreover, the variational design with explicit prior and likelihood functions offers a flexible plug-and-play framework that can be easily customized to suit various protein design tasks.
APA
Bogensperger, L., Narnhofer, D., Allam, A., Schindler, K. & Krauthammer, M.. (2025). A Variational Perspective on Generative Protein Fitness Optimization. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:4700-4712 Available from https://proceedings.mlr.press/v267/bogensperger25a.html.

Related Material