Related Experiment Video
Updated: Jul 12, 2026
09:25
NMR 15N Relaxation Experiments for the Investigation of Picosecond to Nanoseconds Structural Dynamics of Proteins
Published on: November 1, 2024
Learning millisecond protein dynamics from what is missing in NMR spectra
Hannah K Wayment-Steele1, Gina El Nesr2, Ramith Hettiarachchi3,4
1Department of Integrated Structural and Computational Biology, Scripps Research & Howard Hughes Medical Institute, La Jolla, CA.
Biorxiv : the Preprint Server for Biology
|July 10, 2026
Summary
Researchers developed a deep learning model, Dyna-1, to predict protein dynamics from missing Nuclear Magnetic Resonance (NMR) data. This advances understanding of protein function by analyzing micro-to-millisecond motions.
Area of Science:
- Biophysics
- Structural Biology
- Computational Biology
Background:
- Protein function is intrinsically linked to conformational changes occurring on micro- to millisecond (µs-ms) timescales.
- A significant gap exists in standardized, large-scale experimental data for characterizing these protein dynamics.
- Nuclear Magnetic Resonance (NMR) spectroscopy can reveal µs-ms dynamics, but these motions often lead to signal broadening and unassigned residues in spectral data.
Purpose of the Study:
- To develop a predictive model for understanding protein dynamics based on commonly observed NMR data.
- To leverage deep learning to interpret missing residue assignments in NMR spectra as indicators of µs-ms dynamics.
- To establish a link between protein dynamics, sequence conservation, and biological function.
Main Methods:
- Curated over 100 Nuclear Magnetic Resonance (NMR) relaxation datasets.
- Utilized deep learning models to predict missing residue assignments in protein NMR spectra.
- Trained models to predict exchange broadening, a signature of µs-ms dynamics.
- Integrated the multimodal language model ESM-3 into the best-performing model, Dyna-1.
Main Results:
- Deep learning models successfully predicted missing NMR assignments, correlating with µs-ms dynamics.
- The Dyna-1 model, incorporating ESM-3, accurately predicted exchange broadening.
- Dyna-1 demonstrated superior prediction of dynamics crucial for biological functions like enzyme catalysis and ligand binding.
- Residues exhibiting µs-ms exchange dynamics were found to be more conserved across protein sequences.
Conclusions:
- Missing NMR assignments can serve as a proxy for unobserved µs-ms protein dynamics.
- Deep learning, particularly models like Dyna-1, can effectively predict and analyze these dynamics.
- The study provides a novel approach to link protein dynamics to function and conservation.
- The developed datasets and models offer a transformative resource for studying protein dynamics and function.
Related Concept Videos
01:02¹H NMR: Interpreting Distorted and Overlapping Signals
Spin systems where the difference in chemical shifts of the coupled nuclei is greater than ten times J are called first-order spin systems. These nuclei are weakly coupled, and their chemical shifts and coupling constant can generally be estimated from the well-separated signals in the spectrum.
As Δν decreases and the signals move closer, the doublets appear increasingly distorted. The intensities of the inner lines increase at the cost of those of the outer lines as the signals are slanted or...
As Δν decreases and the signals move closer, the doublets appear increasingly distorted. The intensities of the inner lines increase at the cost of those of the outer lines as the signals are slanted or...
01:14NMR Spectrometers: Resolution and Error Correction
When magnetic nuclei in a sample achieve resonance and undergo relaxation, the signal detected in NMR is an approximately exponential free induction decay. Fourier transform of an exponential decay yields a Lorentzian peak in the frequency domain. Lorentzian peaks in an NMR spectrum are defined by their amplitude, full width at half maximum, and position, where the peak width is governed by the spin-spin relaxation time alone. In real experiments, however, the applied magnetic field is rendered...
00:52¹H NMR of Conformationally Flexible Molecules: Temporal Resolution
At room temperature, the chair conformer of cyclohexane undergoes rapid ring flipping between two equivalent chair conformers at a rate of approximately 105 times per second. These two chair conformers are in equilibrium. The rapid ring flipping results in the interconversion of the axial proton to an equatorial proton and an equatorial to the axial proton. Such interconversions are too rapid and cannot be detected on the NMR timescale. Hence, the NMR spectrometer cannot distinguish between the...
01:04¹³C NMR: ¹H–¹³C Decoupling
The probability of having two carbon-13 atoms next to each other is negligible because of the low natural abundance of carbon-13. Consequently, peak splitting due to carbon-carbon spin-spin coupling is not observed in spectra. However, protons up to three sigma bonds away split the carbon signal according to the n+1 rule, resulting in complicated spectra.
A broadband decoupling technique is used to simplify these complex, sometimes overlapping, signals. Broadband decoupling relies on a...
A broadband decoupling technique is used to simplify these complex, sometimes overlapping, signals. Broadband decoupling relies on a...
01:10¹H NMR of Labile Protons: Temporal Resolution
Protons bonded to heteroatoms such as nitrogen and oxygen exhibit a range of chemical shift values. This is due to the varying degree of hydrogen bonding between the proton and the heteroatom in other molecules. The extent of hydrogen bonding affects the electron density around the proton, thereby giving different chemical shift values for the protons in the proton NMR spectrum.
The –OH proton in alcohols typically appears in the range of δ 2 to 5 ppm but can vary depending on the specific...
The –OH proton in alcohols typically appears in the range of δ 2 to 5 ppm but can vary depending on the specific...
01:28Atomic Nuclei: Types of Nuclear Relaxation
Nuclear relaxation restores the equilibrium population imbalance and can occur via spin–lattice or spin–spin mechanisms, which are first-order exponential decay processes.
In spin–lattice or longitudinal relaxation, the excited spins exchange energy with the surrounding lattice as they return to the lower energy level. Among several mechanisms that contribute to spin–lattice relaxation, magnetic dipolar interactions are significant. Here, the excited nucleus transfers energy to a nearby...
In spin–lattice or longitudinal relaxation, the excited spins exchange energy with the surrounding lattice as they return to the lower energy level. Among several mechanisms that contribute to spin–lattice relaxation, magnetic dipolar interactions are significant. Here, the excited nucleus transfers energy to a nearby...