Sampling alternative conformational states of transporters and receptors with AlphaFold2

By Diego del Alamo, Davide Sala, Hassane S. Mchaourab, Jens Meiler. 2022

409 citations recorded by the source corpus.

Equilibrium fluctuations and triggered conformational changes often underlie the functional cycles of membrane proteins. For example, transporters mediate the passage of molecules across cell membranes by alternating between inward- and outward-facing states, while receptors undergo intracellular structural rearrangements that initiate signaling cascades. Although the conformational plasticity of these proteins has historically posed a challenge for traditional de novo protein structure prediction pipelines, the recent success of AlphaFold2 (AF2) in CASP14 culminated in the modeling of a transporter in multiple conformations to high accuracy. Given that AF2 was designed to predict static structures of proteins, it remains unclear if this result represents an underexplored capability to accurately predict multiple conformations and/or structural heterogeneity. Here, we present an approach to drive AF2 to sample alternative conformations of topologically diverse transporters and G-protein-coupled receptors that are absent from the AF2 training set. Whereas models of most proteins generated using the default AF2 pipeline are conformationally homogeneous and nearly identical to one another, reducing the depth of the input multiple sequence alignments by stochastic subsampling led to the generation of accurate models in multiple conformations. In our benchmark, these conformations spanned the range between two experimental structures of interest, with models at the extremes of these conformational distributions observed to be among the most accurate (average template modeling score of 0.94). These results suggest a straightforward approach to identifying native-like alternative states, while also highlighting the need for the next generation of deep learning algorithms to be designed to predict ensembles of biophysically relevant states.

Subfields: Molecular Biology, Computational Biology & Bioinformatics, Computational Chemistry & Drug Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

In Computational Biology & Bioinformatics

In Computational Chemistry & Drug Discovery

Citations in the corpus

Predicting multiple conformations via sequence clustering and AlphaFold2 (2023, positive): “Despite these demonstrations of shortcomings, it was shown that subsampling the input MSA enables AF2 to predict known conformational changes of transporters 14 .”
State-specific protein–ligand complex structure prediction with a multiscale deep generative model (2022, positive): “During model training, an MSA bootstrapping technique 73 was applied to sample diverse template structures from AF2 as auxiliary model inputs complementary to alternative experimentally determined conformations of the same proteins retrieved from the PDB (‘Datasets and training summary’).”
High-resolution de novo structure prediction from primary sequence (2022, background): “…lines will continue to enable a wide spectrum of protein science applications, such as multi-state conformational sampling, where AlphaFold2 has had some success to regions for which predictions are challenging ( 41 ), variant effect prediction, protein-protein interactions and protein docking.”
Improved multimer prediction using massive sampling with AlphaFold in CASP15 (2023, methodology): “8 Alternative methods to achieve greater diversity among generated models include increasing the number of times the prediction is recycled in the network, 9 randomly perturbing, 10 or altering the input MSA.”
Estimating conformational heterogeneity of tryptophan synthase with a template‐based Alphafold2 approach (2022, methodology): “In recent papers, different strategies have been proposed to overcome this static picture provided by AF2: (a) del Alamo et al. 27 the number of sequences of the MSA provided to AF2 was modified to contain as few as 16 sequences, they also reduced the number of recycles to 1, and avoid the final MD ”
From sequence to Boltzmann weighted ensemble of structures with AlphaFold2-RAVE (2022, background): “Email: ptiwary@umd.edu partly solved with the simple realization that reducing the size of the input MSA increases the conformational diversity explored in AF2 3,6 .”