A brief chronological list of papers behind the thinking that led to the current state of LSP-MD: the experimental and theoretical work that established the nature and the role of conformational entropy, and the methodological work that has followed.
The theoretical foundation. Showed that an allosteric signal can be transmitted through changes in thermal vibrations alone, with no rearrangement of the average structure. Vibrational modes are excitable at room temperature, and ligand binding can shift the entropy associated with them between functional states.
Linked NMR-derived order parameters to free-energy estimates and noted that ignoring correlations between motions overestimates entropy. The first quantitative warning that per-residue measurements miss the collective character of the variable they are trying to estimate.
First application of LSP alignment to a real biological problem. The hydrophobic spines of the protein kinase fold, now a textbook feature of kinase activation, emerged from the method's pattern-matching across multiple structures.
Truncating a seven-residue C-terminal helix in PDZ3 reduces peptide-binding affinity over twentyfold while leaving the crystal structure unchanged. The cleanest experimental example of a system that structure cannot explain. The system is addressed computationally by the LSP-MD work below.
Demonstrated experimentally that a regulatory protein can switch between functional states almost entirely through changes in conformational entropy, with negligible structural rearrangement. The Cooper–Dryden prediction observed in vivo, twenty-eight years later.
The violin model articulated. Regulation of activity is not transmitted through a domino chain from one site to another, but achieved through collective sidechain dynamics across the whole protein.
Comprehensive review of the experimental program for measuring conformational entropy by NMR: the entropy meter framework, the calibration data behind it, and what a decade of measurements across diverse protein systems has revealed about the role of sidechain dynamics in molecular recognition. The reference work for the experimental side of the field.
Centralities computed on LSP-derived networks for protein kinase A identified residues already known to be functionally important. The first application of LSP to molecular dynamics data, with results that converged within 10 ns trajectories.
First application of LSP-derived networks to an allosteric system. The cAMP-bound and unbound states of CAP produced visibly distinct sidechain networks that organized themselves into the protein's known structural and functional units, without any input about three-dimensional geometry.
Reviews the contemporary picture of how conformational entropy contributes to protein function: distinguishing distinct mechanisms including entropic prepayment, spatial redistribution, and entropy-supported catalytic ensembles. The current statement of where the field has arrived in interpreting the variable.
Method paper for LSP-MD. Establishes the physical basis of the approach in sub-picosecond thermal fluctuations of sidechains, links the resulting networks back to the Cooper–Dryden theoretical model, and adapts LSP for analysis of MD trajectories.
PDZ3 revisited with LSP-MD. Time-resolved community analysis reveals that sidechain network organization reorganizes stochastically on the 50–200 ns timescale (entropy sloshing), and that PCA cleanly separates this background from the genuine allosteric signal between the binding groove and the C-terminal helix.