Napiorkowska, M. | Search bioRxiv

bioRxiv2024

ProteusAI: An Open-Source and User-Friendly Platform for Machine Learning-Guided Protein Design and Engineering

AO_SCPLOWBSTRACTC_SCPLOWProtein design and engineering are crucial for advancements in biotechnology, medicine, and sustainability. Machine learning (ML) models are used to design or enhance protein properties such as stability, catalytic activity, and selectivity. However, many existing ML tools require specialized expertise or lack open-source availability, limiting broader use and further development. To address this, we developed ProteusAI, a user-friendly and open-source ML platform to streamline protein engineering and design tasks. ProteusAI offers modules to support researchers in various stages of the design-build-test-learn (DBTL) cycle, including protein discovery, structure-based design, zero-shot predictions, and ML-guided directed evolution (MLDE). Our benchmarking results demonstrate ProteusAIs efficiency in improving proteins and enyzmes within a few DBTL-cycle iterations. ProteusAI democratizes access to ML-guided protein engineering and is freely available for academic and commercial use. Future work aims to expand and integrate novel methods in computational protein and enzyme design to further develop ProteusAI.

bioinformatics↗

bioRxiv2023

YeastIT: Reducing mutational bias for in vivo directed evolution using a novel yeast mutator strain based on dual adenine-/cytosine-targeting and error-prone DNA repair

Engineering proteins with new functions and properties often requires navigating large sequence spaces through rounds of iterative improvement. However, a disparity exists between the gradual pace of natural long-term evolution and a typical laboratory evolution workflow that relies on enriching functional variants from highly diverse in vitro generated libraries through very few screening rounds. Laboratory experiments often eschew presumed natural strategies such as neutral/non-adaptive and multi-phase evolution trajectories, and therefore mutagenesis technologies suitable for long nature-like timescales are needed. Here, we introduce YeastIT, a novel in vivo mutagenesis tool for protein engineering that leverages an S. cerevisiae strain engineered to exhibit mutagenic activity directed to the gene of interest, allowing its continuous diversification. Mutagenesis is achieved by generating DNA damage through nucleoside deamination, followed by introduction of mutations by harnessing the process of error-prone DNA translesion synthesis. By eliminating the transformation step, YeastIT allows multiple rounds of screening or selection without interruptions for library diversification, thereby enabling long-term and continuous evolution campaigns. Our characterization of the mutational spectrum and frequency of the YeastIT-generated libraries, and its comparison to other methods (error-prone PCR, PACE, MutaT7, eMutaT7, OrthoRep, TRIDENT, EvolVR) demonstrates comparable mutation rates combined with a significant reduction in mutagenic bias relative to most of the alternatives. To validate YeastIT, we carried out directed evolution of a DARPin binding protein to achieve a 15-fold improved affinity. YeastIT thus provides a tool for exploring different evolutionary trajectories which overcomes previous limitations of variant availability (due to bias and low mutation rates) and emulates the way proteins emerge in Nature.

synthetic biology↗