Beta release for use and testing purposes. Subject to changes and additional feature releases. Expect downtimes during updates.

Update in progress (features added stepwise, please be patient while the interface and servers are updated)

This is a beta release for the Enzymares software toolbox – enzyme bioprospecting. We take security and model correctness/accuracy very seriously; therefore, updates are implemented gradually. Please note that delays may occur.

Enzymares Toolbox

Bridging Data and Machine Learning for Enzyme Innovation & Selection

The Enzymares Toolbox is a VLAIO-funded, open-source platform developed through a collaborative initiative. Designed to accelerate enzyme discovery and engineering, this toolbox integrates computational modelling with curated biological data to empower researchers in academia and industry. See https://www.ugent.be/marine/en/partnerships/enzymares for more information.

Core Capabilities

Curated Enzyme Database

Aggregates enzyme data from public repositories (e.g., BRENDA, UniProt, Rhea, OBIS, Copernicus,...), including environmental samples from diverse organisms. This repository enables identification of candidate enzymes for specific applications based on experimental and environmental data.

Proprietary In-House Models

Enzymares platform leverages proprietary in-house machine learning models for enhanced predictive accuracy, optimised with expanded curated datasets. Our models have key advances in prediction of enzyme temperature, pH, and enzyme classification.

Integration of state-of-the-art models

Leverages combined pipelines integrating state-of-the-art (SOTA) machine learning models to predict enzyme properties, such as stability, activity, and substrate specificity. This platform offers ensemble approaches for several SOTA models, carefully curated to only include promising and powerful models.

Genomic and Proteomic context

Includes deep learning annotation, experimental and similarity information, such as sequence and/or structure similarity for improved accuracy.

Key Features

Unified Prediction Pipelines

Seamlessly combines multiple SOTA models for simplified, high-confidence predictions.

Environmental Data Insights

Proprietary computationally annotated datasets for extremophiles and industrial strains expand the search space for novel enzyme candidates.

Modular Architecture

Customize workflows for specific enzymes or research goals without compromising scalability.

Cross-Platform Accessibility

Deploy locally or via cloud-based solutions using Docker/Singularity containers.

Applications

Candidate Screening

Identify enzymes with desired characteristics from environmental or synthetic databases.

Enzyme Optimization

Predict the impact of mutations on catalytic efficiency or thermostability.

Developed under the VLAIO funding program, the Enzymares Toolbox is freely available at enzymares.org. Join a global community leveraging integrated data and cutting-edge modelling to redefine enzyme innovation.