PymolFold: A PyMOL Plugin for API-driven Structure Prediction and Quality Assessment

Abstract

Deep learning has transformed protein structure prediction, yet many experimental scientists face barriers in accessing state-of-the-art (SOTA) models due to technical complexity and hardware requirements. To address this, we present PymolFold, an open-source PyMOL plugin that seamlessly integrates cutting edge API-based protein structure predictors such as ESM-3 and Boltz2 into the molecular visualization environment. PymolFold supports both graphical and command-line interfaces for flexible usage and incorporates PXMeter, an open-source Python package for quantitative evaluation of protein structure predictions against reference data. Together, these features establish a unified "predict-visualize-analyze" workflow, lowering technical entry barriers and broadening access to advanced structural modeling. PymolFold is freely available at https://github.com/jinyuansun/PymolFold.