Evaluating the electron density model by applying an imaginary modification

TL;DR

This paper introduces an evaluation function for electron density models that applies an imaginary modification and estimates the model's accuracy, showing promise at high-resolution data below 2.5 Å.

Contribution

It proposes a novel evaluation method using imaginary electron density modifications and estimated structure amplitudes to assess model accuracy without needing the correct model.

Findings

Evaluation function is potentially valid at resolutions better than 2.5 Å.

The method provides a way to estimate model quality without the true model.

Limited tests on simulated data support the approach's feasibility.

Abstract

A function has been proposed to evaluate the electron density model constructed by inverse Fourier transform using the observed structure amplitudes and trial phase set. The strategy of this function is applying an imaginary electron density modification to the model, and then measuring how well the calculated structure amplitudes of the modified model matches the expected structure amplitudes for the modified correct model. Since the correct model is not available in advance, a method has been developed to estimate the structure amplitudes of the modified correct model. With the estimated structure amplitudes of the modified correct model, the evaluation function can be calculated approximately. Limited tests on simulated diffraction data indicate that this evaluation function may be valid at the data resolution better than 2.5 {\AA}.