About Mobius

Mobius provides a robust and versatile Bayesian optimization (BO) approach, designed to aid in the design and optimization of peptides within a fully automated, closed-loop Design-Make-Test (DMT) pipeline context.

Equipped to manage an array of peptide scaffolds, including complex structures such as macrocycles, as well as non-natural amino acids using the HELM notation, Mobius employs a variety of combinatorial strategies. These strategies, which include alanine, random, homolog, and property-based scanning, facilitate the generation of an initial batch of peptides, enabling you to jumpstart your optimization project. Moreover, Mobius features an array of fingerprint methods that are adaptable to a wide spectrum of peptide types.

Mobius shines in its flexibility and modularity, allowing for seamless integration with custom surrogate models, fingerprint methods, and filters. This ensures that Mobius can be tailored to meet the specific demands and requirements of your peptide optimization projects.

The source code is available under the Apache license at https://git.scicore.unibas.ch/schwede/mobius.

The method is described in the following paper: J. Eberhardt, A. Lees, M. Lill, T. Schwede. (2024). Combining Bayesian optimization with sequence- or structure-based strategies for optimization of peptide-binding protein.