rFactors

RFactors, or R-factors, are a family of reliability metrics used in X-ray crystallography and related structural methods to quantify how well a crystallographic model agrees with observed diffraction data. The most common form, the R-factor, measures the discrepancy between observed structure-factor amplitudes (Fo) and those calculated from the model (Fc). It is typically defined as R = sum over reflections of |Fo − Fc| divided by sum over reflections of |Fo|. Lower R-factors indicate better agreement, but interpretation depends on data quality, resolution, and refinement strategy.

Two widely reported variants are R-work and R-free. R-work is calculated using the reflections included in the refinement process, while R-free uses a separate subset of reflections not used in refinement to provide an unbiased check against overfitting. R-free is generally considered a more robust indicator of model validity, especially in the presence of limited data or complex models.

R-factors are complemented by other validation metrics. Examples include R-merge and related statistics that describe data redundancy and consistency, as well as map quality indicators and geometry-based checks (such as bond lengths, angles, and Ramachandran statistics) to assess model plausibility beyond crystallographic fit.

Limitations and interpretation considerations are important. R-factors depend on data quality, resolution, completeness, and model complexity; a low R-factor does not guarantee a correct or unique structure, and a high R-factor can sometimes reflect data limitations rather than a fundamentally flawed model. Consequently, researchers use R-factors alongside complementary refinements and validation tools to judge structural reliability.