Brookhaven National Laboratory has been instrumental in the development of ESMBind, an innovative tool for protein discovery and characterization. In particular, it focuses on the interactions of metal binding proteins. Learn more about how AI scientist Xin Dai and structural biologist Qun Liu developed ESMBind. Underneath, it runs advanced foundation models from Meta, namely ESM-IF and ESM-2. This innovative screening tool not only predicts the three-dimensional shapes of proteins but enhances understanding of their functional interactions with metals like zinc.
ESMBind is very powerful, though. It has the capacity to run hundreds of thousands of simulations per day and beyond all other AI models, excels at precisely predicting protein structures and functions. There are dire consequences far beyond the ivory tower if this research continues to be ignored. The project’s objectives are to protect biofuel crops from infectious diseases and develop sustainable agricultural practices.
Groundbreaking Technology Behind ESMBind
The creation of ESMBind relies heavily on advanced computational methods that use information gained from data on existing protein structures and sequences. Nearly all of the training data for ESMBind is derived from X-ray crystallography experiments performed at NSLS-II and other synchrotron facilities. This strong background makes possible ESMBind for modeling unknown proteins, predicting their structures and possible functions.
By focusing on proteins related to metal interactions, researchers can explore deeper into understanding vital processes that affect plant health and crop resilience.
“ESMBind is a screening tool to find proteins that bind to the metals of interest.”
According to research, there is an important opportunity. ESMBind could open the floodgates to discover candidate proteins that traffic pathogenicity in crop infection, including sorghum, an important biofuel crop. Through successful application of ESMBind, Liu and Dai were able to predict the function of the shape-shifting proteins involved with Na-straw. This fungus is a major biotic stress to sorghum crops.
Liu states:
Implications for Agriculture and Biofuels
This approach aligns with the growing need for sustainable agricultural practices that ensure food security while promoting alternative energy sources. That means researchers can increase biofuel production by leveraging land that can’t support traditional crops. This strategy helps them expand production of fuel without putting food at risk.
“We do not want biofuel crops to compete with crops for food. Instead, we need to grow these bioenergy plants on nutritionally deficient land.”
Liu highlights the urgency of protecting valuable crops from diseases:
Given increasing global demand for biofuels, this commitment to protecting other crops is absolutely indispensable.
“Protecting plants and biofuel crops from infectious diseases is a research priority for the plant sciences group within the Brookhaven Lab Biology Department.”
With ESMBind, we are taking a big step in agricultural biotechnology by using new machine learning methods to make protein modeling more efficient and powerful. Liu believes that such technology can accelerate research efforts:
Enhancing Research Through Machine Learning
ESMBind has wider applications that promise to go well beyond disease resistance. It can aid engineers in developing bioengineered microbes that more efficiently extract rare earth elements. Liu envisions a future where designed proteins can selectively capture these elements, facilitating their recovery from environmental sources:
“We believe there’s opportunity to leverage machine learning, a form of AI, to speed up the creation of useful protein models.”
This unique, out-of-the-box project really showcases the versatility of ESMBind. It highlights the urgent need to incorporate AI technology into biological research.
“If we can design a protein to fold and capture a rare earth element in a specific way, we might be able to engineer microbes to make that protein and use them to extract and recover that critical mineral.”
This innovative approach not only showcases the versatility of ESMBind but also reaffirms the importance of integrating AI technology into biological research.