analiza aiqbench
This research introduces sequence kernels leveraging evolutionary substitution matrices and local linearity to build data-efficient Gaussian processes for protein property prediction. The approach notably outperforms foundation model embeddings and local supervised methods by effectively incorporating structural information for multi-task learning. For the machine learning community, this offers a robust alternative to heavy pre-trained models when working with sparse experimental datasets in protein design.
Flexible Kernels for Protein Property Prediction
arXiv:2606.11057v1 Announce Type: new Abstract: Despite its importance to applications in protein design, predicting protein properties like binding affinity and thermostability from sparse experimental data remains a significant challenge. Accordingly, we introduce a class of sequence kernels that exploit evolutionary substitution matrices as well as local linearity and demonstrate that the resulting Gaussian processes provide data-efficient models of protein property landscapes, frequently outperforming alternatives that rely on foundation model embeddings. Furthermore--by learning what are in effect structure-aware substitution matrices--we show that our kernels can readily incorporate structural information from foundation models. We demonstrate that these structure-conditioned kernels are well suited to multi-task learning across multiple protein property landscapes and can decisively outperform local supervised learning methods.
Abstract:Despite its importance to applications in protein design, predicting protein properties like binding affinity and thermostability from sparse experimental data remains a significant challenge. Accordingly, we introduce a class of sequence kernels that exploit evolutionary substitution matrices as well as local linearity and demonstrate that the resulting Gaussian processes provide data-efficient models of protein property landscapes, frequently outperforming alternatives that rely on foundation model embeddings. Furthermore--by learning what are in effect structure-aware substitution matrices--we show that our kernels can readily incorporate structural information from foundation models. We demonstrate that these structure-conditioned kernels are well suited to multi-task learning across multiple protein property landscapes and can decisively outperform local supervised learning methods.
Machine Learning (cs.LG); Biomolecules (q-bio.BM); Machine Learning (stat.ML)
arXiv:2606.11057 [cs.LG]
(or arXiv:2606.11057v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2606.11057 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 210 cuvinte
Subiecte conexe
Advancing AI for biology: Teaching models to design and characterize antibodies
Amazon Science · 21 sept. 2026
Revisiting Thinning Methods for Kernel Learning Problems
arXiv · 9 sept. 2026
PhiFold: Towards Dynamic Protein Design with Physics-Structured Covariance Modeling
arXiv · 29 sept. 2026
Towards A Generative Protein Evolution Machine with DPLM-Evo
arXiv · 5 iun. 2026