Direct Phasing of Nanocrystal Diffraction

TL;DR

This paper introduces a novel direct phasing algorithm for nanocrystal diffraction data that reconstructs particle structures using intensity and gradient information, effective even at low resolution and low signal-to-noise ratios.

Contribution

The method reconstructs nanocrystal particles from intensity gradients without requiring high-resolution data or contrast constraints, unlike traditional approaches.

Findings

Successful reconstruction of lysozyme nanocrystals at low SNR

Effective at low resolution without contrast constraints

Works with simulated data down to SNR of 2

Abstract

Recent experiments at free-electron laser x-ray sources have been able to resolve the intensity distributions about Bragg peaks in nanocrystals of large biomolecules. Information derived from small shifts in the peak positions augment the Bragg samples of the particle intensity with samples of its gradients. Working on the assumption that the nanocrystal is entirely generated by lattice translations of a particle, we develop an algorithm that reconstructs the particle from intensities and intensity gradients. Unlike traditional direct phasing methods that require very high resolution data in order to exploit sparsity of the electron density, our method imposes no constraints on the contrast other than positivity and works well at low resolution. We demonstrate successful reconstructions with simulated P1 lysozyme nanocrystal data down to a signal-to-noise ratio of 2 in the intensity gradients.