Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network

Abstract

Cryptic pockets expand the scope of drug discovery by enabling targeting of proteins currently considered undruggable because they lack pockets in their ground state structures. However, identifying cryptic pockets is labor-intensive and slow. The ability to accurately and rapidly predict if and where cryptic pockets are likely to form from a structure would greatly accelerate the search for druggable pockets. Here, we present PocketMiner, a graph neural network trained to predict where pockets are likely to open in molecular dynamics simulations. Applying PocketMiner to single structures from a newly curated dataset of 39 experimentally confirmed cryptic pockets demonstrates that it accurately identifies cryptic pockets (ROC-AUC: 0.87) >1,000-fold faster than existing methods. We apply PocketMiner across the human proteome and show that predicted pockets open in simulations, suggesting that over half of proteins thought to lack pockets based on available structures likely contain cryptic pockets, vastly expanding the potentially druggable proteome.

Accelerating Cryptic Pocket Discovery Using AlphaFold

“…In those cases, AF predicts a structure with less than 1.2 Å root-mean-square deviation (RMSD) to the holo structure in the cryptic site (i.e., using all heavy atoms within 5 Å of where the cryptic ligand binds for the RMSD calculation). Interestingly, AlphaFold generates open states of the Niemann-Pick C2 Protein that were not discovered in 2 μs of adaptive sampling simulations (Figure B) . However, AlphaFold’s ensemble of TEM β-lactamase structures does not include any open states where the Horn or omega pockets are open (Figure S1).…”

“…This result makes PM-II an exception to a general trend. We have found that many cryptic pockets can be discovered with a handful of simulations of intermediate length (i.e., 40 ns) . Furthermore, significant progress has been made in developing algorithms for cryptic pocket discovery, including Markov state models, , enhanced sampling strategies like SWISH, or adaptive sampling approaches like FAST .…”

“…Cryptic pockets, or pockets absent in ligand-free experimental structures, are a promising means to expand the scope of drug discovery. By one estimate, almost half of all structured domains lack obvious pockets in their experimental structures . These proteins have often been considered “undruggable”.…”

“…While this process has revealed cryptic pockets, it requires knowledge of a ligand a priori . Molecular dynamics simulations can reveal excited states with cryptic pockets that can then be used for structure-based drug design. , However, in certain cases, cryptic pockets may not be discovered by simulations because cryptic pocket opening motions may be slow (e.g., Niemann-Pick C2 Protein in Meller et al). Two classes of slow motions include side chain ring flipping events and secondary structure rearrangements which can both occur on microsecond and slower time scales.…”

Discovery of a cryptic pocket in the AI-predicted structure of PPM1D phosphatase explains the binding site and potency of its allosteric inhibitors

“…Next, inspired by recent success in capturing cryptic pocket formation in molecular dynamics simulations, ( Hollingsworth et al, 2019 ; Sztain et al, 2021 ; Zimmerman et al, 2021 ; Cruz et al, 2022 ; Meller et al, 2023b ; Meller et al, 2023c ), we tested whether simulations launched from the AF structure could reveal cryptic pockets that encompass the flap or the hinge. We used an adaptive sampling algorithm FAST ( Zimmerman and Bowman, 2015 ) to search for cryptic pockets.…”