# NMR Spectral Alignment Utilizing a CryoEM Motion Correction Algorithm

## TL;DR

This paper introduces ANSA, a new tool that uses cryoEM motion correction to automatically align NMR spectra, improving accuracy and reducing human bias.

## Contribution

The novel contribution is adapting cryoEM motion correction algorithms for NMR spectral alignment, enabling objective and automated processing.

## Key findings

- ANSA improves cross-correlation scores from 0.33 to 1.00 in controlled tests.
- The algorithm successfully aligns spectra across varying experimental conditions and works with 2D and 3D datasets.
- ANSA achieves 0.96 correlation in real-world applications with previously misaligned spectra.

## Abstract

With recent advances in magic-angle spinning (MAS) solid-state
NMR (SSNMR) resolution, precise spectral alignment has become a critical
bottleneck in data processing workflows. While solution NMR employs
deuterium lock systems, most SSNMR probes still lack this capability;
though a lock corrects for magnet drift and instabilities, it is not
alone sufficient to account for field gradients, sample temperature
differences, and pulse sequence effects that can contribute to referencing
errors among several data sets. These offsets become particularly
problematic in the lengthy multidimensional experiments that provide
the foundation for resonance assignment and structure determination
procedures. Currently, researchers rely on manual alignment through
visual peak inspectiona qualitative approach that often overemphasizes
prominent, outlying peaks while overlooking subtle, global patterns.
This subjective process becomes increasingly impractical for use cases
with lower sensitivity, such as large proteins with thousands of peaks.
To address these challenges, here we present Automated NMR Spectral
Alignment (ANSA), a program that adapts cryo-electron
microscopy motion correction principles to NMR spectroscopy. ANSA treats NMR spectra as images and applies cross-correlation
functions to determine optimal alignment, improving cross-correlation
scores from 0.33 to 1.00 in controlled tests and achieving 0.96 correlation
in real-world applications with previously misaligned spectra. The
algorithm successfully aligns spectra across varying experimental
conditions, corrects shifts in long-duration experiments, and works
with 2D and 3D data sets, with approaches that can be readily extended
to additional dimensions. By eliminating human bias and providing
objective, consistent spectral alignment, ANSA enhances
scientific rigor, improves reproducibility between experiments, and
enables automation of critical data processing steps. The software
is freely available as an open-source tool, ready for integration
into existing NMR workflows.