Quality assessment of protein structures is an important part of experimental structure validation and plays a crucial role in protein structure prediction. Most current scoring functions are primarily de-signed to rank alternative models of the same sequence supporting model selection whereas the prediction of the absolute quality of an individual protein model has received little attention in the field. However, reliable absolute quality estimates are crucial to assess the suitability of a model for specific biomedical applications. We have developed a new absolute measure for the quality of protein models, which provides an estimate of the “degree of nativeness” of the structural features observed in a model and describes the likelihood that a given model is of comparable quality to experimental structures. Model quality scores for individual models are expressed as “Z-scores” in comparison to scores obtained for high-resolution crystal structures. We demonstrated the ability of the newly introduced QMEAN Z-score to detect experimentally solved protein structures containing significant errors, as well as to evaluate theoretical protein models.

Scoring functions can be broadly categorized into two groups: (1) approaches being able to estimate the quality of a single model without relying on consensus information and (2) consensus or clustering methods relying on the comparative analysis of the structural similarity among the models in an ensemble. In the course of protein structure prediction typically a set of alternative models is produced, from which the most accurate candidate is selected using a scoring function. Estimating the quality of protein structure models is a critical step in protein structure prediction because it is the quality of a model which determines its suitability for real-world applications. In general, consensus methods perform significantly better in assessing server models than physics-based or evolutionary methods operating on single models. Nevertheless, the latter category of methods plays an important role in assessing individual or small sets of models. Hybrid methods combining single model scoring functions with structural consensus information can be used to counteract some of the shortcomings of pure consensus methods. The idea of using a scoring function such as QMEAN operating on single models in order to prioritize the contribution of the models within the ensemble used to generate the consensus score has been shown to be a promising strategy both for the global and especially for the local quality estimation of models.