Interpretability Driven Evolutionary Approach for the Design of Biological Sequences

Akash Pandey, Wei Chen, Sinan Keten
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:96194-96211, 2026.

Abstract

Designing biological sequences such as proteins and DNA for desired properties is challenging due to vast search spaces and limited wet lab evaluation budgets. Current evolutionary approaches ignore sequential dependencies and rely on random mutations, which scale poorly for long sequences. In contrast, reinforcement learning (RL) and generative models explicitly model sequence structure but require large datasets to guide generation toward the target properties. These limitations suggest the need for a method that combines the sample efficiency of evolutionary approaches with the ability to exploit sequential structure. In this work, we propose a novel evolutionary approach, $\texttt{IDEAS}$, in which mutations are guided by an explainable model. The model identifies critical motifs in high-fitness sequences and uses them to mutate non-critical positions. Across eight continuous-property datasets, seven baselines, and three evaluation budgets, $\texttt{IDEAS}$ achieves a 19% acceleration in design while maintaining a favorable position on the Pareto curve balancing acceleration, diversity, and novelty.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-pandey26a, title = {Interpretability Driven Evolutionary Approach for the Design of Biological Sequences}, author = {Pandey, Akash and Chen, Wei and Keten, Sinan}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {96194--96211}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/pandey26a/pandey26a.pdf}, url = {https://proceedings.mlr.press/v306/pandey26a.html}, abstract = {Designing biological sequences such as proteins and DNA for desired properties is challenging due to vast search spaces and limited wet lab evaluation budgets. Current evolutionary approaches ignore sequential dependencies and rely on random mutations, which scale poorly for long sequences. In contrast, reinforcement learning (RL) and generative models explicitly model sequence structure but require large datasets to guide generation toward the target properties. These limitations suggest the need for a method that combines the sample efficiency of evolutionary approaches with the ability to exploit sequential structure. In this work, we propose a novel evolutionary approach, $\texttt{IDEAS}$, in which mutations are guided by an explainable model. The model identifies critical motifs in high-fitness sequences and uses them to mutate non-critical positions. Across eight continuous-property datasets, seven baselines, and three evaluation budgets, $\texttt{IDEAS}$ achieves a 19% acceleration in design while maintaining a favorable position on the Pareto curve balancing acceleration, diversity, and novelty.} }
Endnote
%0 Conference Paper %T Interpretability Driven Evolutionary Approach for the Design of Biological Sequences %A Akash Pandey %A Wei Chen %A Sinan Keten %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-pandey26a %I PMLR %P 96194--96211 %U https://proceedings.mlr.press/v306/pandey26a.html %V 306 %X Designing biological sequences such as proteins and DNA for desired properties is challenging due to vast search spaces and limited wet lab evaluation budgets. Current evolutionary approaches ignore sequential dependencies and rely on random mutations, which scale poorly for long sequences. In contrast, reinforcement learning (RL) and generative models explicitly model sequence structure but require large datasets to guide generation toward the target properties. These limitations suggest the need for a method that combines the sample efficiency of evolutionary approaches with the ability to exploit sequential structure. In this work, we propose a novel evolutionary approach, $\texttt{IDEAS}$, in which mutations are guided by an explainable model. The model identifies critical motifs in high-fitness sequences and uses them to mutate non-critical positions. Across eight continuous-property datasets, seven baselines, and three evaluation budgets, $\texttt{IDEAS}$ achieves a 19% acceleration in design while maintaining a favorable position on the Pareto curve balancing acceleration, diversity, and novelty.
APA
Pandey, A., Chen, W. & Keten, S.. (2026). Interpretability Driven Evolutionary Approach for the Design of Biological Sequences. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:96194-96211 Available from https://proceedings.mlr.press/v306/pandey26a.html.

Related Material