Fluorescent protein quantum yield prediction has been advanced through the development of a novel algorithm that focuses on the mature chromophore and its three-dimensional microenvironment. This approach recognizes that quantum yield is not solely determined by sequence identity, but rather by the complex interplay between the chromophore and its surrounding environment. The algorithm utilizes a chromophore-centred mechanism graph to model local physical signals and their impact on specific chromophore regions, providing a more nuanced understanding of quantum yield dynamics. By capturing the intricate relationships between chromophore structure and environment, this method has the potential to improve the accuracy of quantum yield predictions1. This matters to researchers and scientists working with fluorescent proteins, as a deeper understanding of quantum yield can inform the design of more efficient and effective fluorescent proteins for various applications.