Bradley, S. A. | bioRxiv Science

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↗

bioRxiv2022

Photoacoustic imaging reveals mechanisms of rapid-acting insulin formulations dynamics at the injection site

ObjectiveUltra-rapid insulin formulations control postprandial hyperglycemia; however, inadequate understanding of injection site absorption mechanisms is limiting further advancement. We used photoacoustic imaging to investigate the injection site dynamics of dye-labeled insulin lispro in the Humalog(R) and Lyumjev(R) formulations using the murine ear cutaneous model and correlated it with results from unlabeled insulin lispro in pig subcutaneous injection model. MethodsWe employed dual-wavelength optical-resolution photoacoustic microscopy to study the absorption and diffusion of the near-infrared dye-labeled insulin lispro in the Humalog and Lyumjev formulations in mouse ears. We mathematically modeled the experimental data to calculate the absorption rate constants and diffusion coefficients. We studied the pharmacokinetics of the unlabeled insulin lispro in both the Humalog and Lyumjev formulations as well as a formulation lacking both the zinc and phenolic preservative in pigs. The association state of insulin lispro in each of the formulations was characterized using SV-AUC and NMR spectroscopy. ResultsThrough experiments using murine and swine models, we show that the hexamer dissociation rate of insulin lispro is not the absorption rate-limiting step. We demonstrated that the excipients in the Lyumjev formulation produce local tissue expansion and speed both insulin diffusion and microvascular absorption. We also show that the diffusion of insulin lispro at the injection site drives its initial absorption; however, the rate at which the insulin lispro crosses the blood vessels is its overall absorption rate-limiting step. ConclusionsThis study provides insights into injection site dynamics of insulin lispro and the impact of formulation excipients. It also demonstrates photoacoustic microscopy as a promising tool for studying protein therapeutics. The results from this study address critical questions around the subcutaneous behavior of insulin lispro and the formulation excipients, which could be useful to make faster and better controlled insulin formulations in the future. HighlightsO_LIHexamer dissociation is not the absorption rate-limiting step for insulin lispro C_LIO_LILyumjev excipients enhance insulin microvascular absorption and diffusion C_LIO_LIVascular endothelial transit determines the overall absorption for insulin lispr C_LIO_LIInsulin diffusion studied for the first time at the injection site of live animals C_LIO_LIIn vivo imaging is a powerful tool to study injection site dynamics C_LI

systems biology↗