Pramodh Vallurupalli

This paper uses deep neural networks to reconstruct 2D NMR spectra from single-pulse experiments, eliminating transfer delays and enabling high-quality methyl correlation maps for large proteins.

This paper develops a deep neural network, FID-Net-2, combined with custom NMR experiments to accurately reconstruct high-resolution aromatic 1H-13C correlation spectra from uniformly 13C-labeled protein samples.

This paper improves the accuracy of exchange parameters in CEST experiments by using K instead of kex to guide B1 field selection, demonstrating better results in two FF domain mutants.