An Investigation of Atomic Structures Derived from X-ray Crystallography and Cryo-Electron Microscopy Using Distal Blocks of Side-Chains

Outlier Profiles of Atomic Structures Derived from X-ray Crystallography and from Cryo-Electron Microscopy

Background: As more protein atomic structures are determined from cryo-electron microscopy (cryo-EM) density maps, validation of such structures is an important task. Methods: We applied a histogram-based outlier score (HBOS) to six sets of cryo-EM atomic structures and five sets of X-ray atomic structures, including one derived from X-ray data with better than 1.5 Å resolution. Cryo-EM data sets contain structures released by December 2016 and those released between 2017 and 2019, derived from resolution ranges 0-4 Å and 4-6 Å respectively. Results: The distribution of HBOS values in five sets of X-ray structures show that HBOS is sensitive distinguishing sets of X-ray structures derived from different resolution ranges-higher than 1.5 Å, 1.5-2.0 Å, 2.0-2.5 Å, 2.5-3.0 Å, and 3.0-3.5 Å. The overall quality of cryo-EM structures is likely improved, as shown in a comparison of cryo-EM structures released before the end of 2016, those between 2017 and 2018, and those between 2018 and 2019. Our investigation shows that leucine (LEU) has a significantly higher rate of HBOS outliers than that of the reference data set (X-ray-1.5) and of other residue types in the cryo-EM data sets. HBOS was able to detect outliers for those residues that are currently marked as green in PDB validation reports. Conclusions: The HBOS profile of a dataset is a potential method to characterize the overall structural quality of the set. Residue LEU deserves special attention since it has a significantly higher HBOS outlier rate in sets of cryo-EM structures and those X-ray structures derived from X-ray data of lower than 2.5 Å resolutions. Most HBOS outlier residues from the EM-0-4-2019 set are located on loops for most types of residues.

A Histogram-based Outlier Profile for Atomic Structures Derived from Cryo-Electron Microscopy

As more atomic structures are determined from cryo-electron microscopy (cryo-EM) density maps, validation of such structures is an important task. We report findings after analyzing the change of cryo-EM structures in a comparison between those released by December 2016 and those released between 2017 and 2019. The cryo-EM models created from density maps with resolution better than 6 Å were divided into six data sets. A histogram-based outlier score (HBOS) was implemented and validation reports were collected from the Protein Data Bank. The results suggest that the overall quality of EM structures released after December 2016 is better than that of structures released before 2017. The conformation qualities of most residue types might have been improved, except for Leucine, Phenylalanine, and Serine in high-resolution datasets (higher than 4 Å). We observe that structures solved from 0-4 Å resolution density maps have an almost identical HBOS profile as that of structures derived from density maps with 4-6 Å resolution.

A Visualization Tool for Cryo-EM Protein Validation with an Unsupervised Machine Learning Model in Chimera Platform

Background: Cryo-electron microscopy (cryo-EM) has become a major technique for protein structure determination. However, due to the low quality of cryo-EM density maps, many protein structures derived from cryo-EM contain outliers introduced during the modeling process. The current protein model validation system lacks identification features for cryo-EM proteins making it not enough to identify outliers in cryo-EM proteins. Methods: This study introduces an efficient unsupervised outlier detection model for validating protein models built from cryo-EM technique. The current model uses a high-resolution X-ray dataset (<1.5 Å) as the reference dataset. The distal block distance, side-chain length, phi, psi, and first chi angle of the residues in the reference dataset are collected and saved as a database of the histogram-based outlier score (HBOS). The HBOS value of the residues in target cryo-EM proteins can be read from this HBOS database. Results: Protein residues with a HBOS value greater than ten are labeled as outliers by default. Four datasets containing proteins derived from cryo-EM density maps were tested with this probabilistic anomaly detection model. Conclusions: According to the proposed model, a visualization assistant tool was designed for Chimera, a protein visualization platform.

A Test Detecting the Outliers for Continuous Distributions Based on the Cumulative Distribution Function of the Data Being Tested

One of the pillars of experimental science is sampling. Based on the analysis of samples, estimations for populations are made. There is an entire science based on sampling. Distribution of the population, of the sample, and the connection among those two (including sampling distribution) provides rich information for any estimation to be made. Distributions are split into two main groups: continuous and discrete. The present study applies to continuous distributions. One of the challenges of sampling is its accuracy, or, in other words, how representative the sample is of the population from which it was drawn. To answer this question, a series of statistics have been developed to measure the agreement between the theoretical (the population) and observed (the sample) distributions. Another challenge, connected to this, is the presence of outliers - regarded here as observations wrongly collected, that is, not belonging to the population subjected to study. To detect outliers, a series of tests have been proposed, but mainly for normal (Gauss) distributions—the most frequently encountered distribution. The present study proposes a statistic (and a test) intended to be used for any continuous distribution to detect outliers by constructing the confidence interval for the extreme value in the sample, at a certain (preselected) risk of being in error, and depending on the sample size. The proposed statistic is operational for known distributions (with a known probability density function) and is also dependent on the statistical parameters of the population—here it is discussed in connection with estimating those parameters by the maximum likelihood estimation method operating on a uniform U(0,1) continuous symmetrical distribution.