High fidelity sampling schedules for NMR spectra of high dynamic range

Revisiting aliasing noise to build more robust sparsity in nonuniform sampling 2D‐NMR

“…Linearization was considered recently with sparse unweighted sampling [8], which actually did not lead to the best artifact reduction in that work. In general, unweighted schedules are prone to poor gap management throughout the schedule, so a small amount of early linearization cannot compensate for the larger problems of unweighted sampling.…”

“…Despite their high potential, these techniques are technically challenging, lack vendor support, and are inaccessible to most. Burst sequences can also red shift sampling noise away from spectral regions of interest [21], while initial uniform regions in NUS have been considered in the evaluation of artifact intensities [8], in power conserving NUS [22], and in triangular backfilling of 2D NUS of 3D NMR [23], which all relate to triangular sampling [24]. This work finds that weak aliasing artifacts from patterned sampling can be a more substantial contribution to sampling noise than previously recognized in sparser NUS and offers a strategy of detecting and remediating these sequences to achieve more robust sparse NUS.…”

“…In 1D-NUS, even 25%-33% sparsity can introduce noise, ridges, and artifactual peaks, which are termed "sampling noise," and the one-dimensional list of samples must solely be robust to all of these pitfalls. Performing more difficult 2D-NMR with NUS and with extended sparsity (25%-33%), from biosolids NMR to natural products work, is receiving more scrutiny [7][8][9][10]. Recent work has focused on developing quantile sampling for challenging 1,1-ADEQUATE experiments [7,11,12] and identifying random seeds for improved Poisson gap sampling [8].…”

“…Performing more difficult 2D-NMR with NUS and with extended sparsity (25%-33%), from biosolids NMR to natural products work, is receiving more scrutiny [7][8][9][10]. Recent work has focused on developing quantile sampling for challenging 1,1-ADEQUATE experiments [7,11,12] and identifying random seeds for improved Poisson gap sampling [8]. Such works have involved extensive testing of sampling schedules that would not be realistic for users to perform as a part of daily spectroscopy.…”

“…Such works have involved extensive testing of sampling schedules that would not be realistic for users to perform as a part of daily spectroscopy. Prior optimizations have also been largely empirical [7,8,13], and there is a need to better understand the barriers to sparser 1D-NUS to help inform schedule design.…”