MS42-P03 The expected log-likelihood gain for decision making in molecular replacement Robert D. Oeffner (Department of Haematology, University of Cambridge, Cambridge, United Kingdom) Pavel V. Afonine (Lawrence Berkeley National Laboratory, Berkeley, United States of America) Claudia Millán (Structural Biology, Molecular Biology Institute of Barcelona, Barcelona, Spain) Massimo Sammito (Structural Biology, Molecular Biology Institute of Barcelona, Barcelona, Spain) Isabel Usón (Structural Biology, Molecular Biology Institute of Barcelona, Barcelona, Spain) Randy J. Read (Department of Haematology, University of Cambridge, Cambridge, United Kingdom) Airlie J. McCoy (Department of Haematology, University of Cambridge, Cambridge, United Kingdom)email: rdo20@cam.ac.ukProtein crystallographers often make assumptions about the solvability of a structure by molecular replacement based on two variables: the sequence identity between the model and target and the resolution of the data. We have recently shown that the solvability of a structure by molecular replacement is, rather, predominantly dependent on four variables: the number of reflections in the data set, the fraction of the scattering for which the model accounts, the RMSD between the model and target, and the measurement errors in the data. Furthermore, the solvability can be quantified with the eLLG (McCoy et al., 2017, Oeffner et al., 2018). The eLLG is the LLGI (Read & McCoy, 2016) expected from a correctly placed model, calculated as a sum of log-likelihoods of each reflection predicted by the model but offset by the sum of log-likelihoods of a model of random atoms. Using the eLLG, the crystallographer can judge whether to pursue molecular replacement or attempt experimental phasing as the quickest path to structure solution. Other applications of the eLLG include determining search order; finding the minimal data requirements for obtaining a molecular replacement solution using a given model; and for decision making in fragment based molecular replacement, in single atom molecular replacement, and for likelihood-guided model pruning.References:

McCoy, A. J., Oeffner, R. D., Wrobel, A. G., Ojala, J. R. M., Tryggvason, K., Lohkamp, B. & Read, R. J. (2017). Proc. Natl. Acad. Sci. U. S. A. 1–5.

Oeffner, R. D., Afonine, P. V., Millán, C., Sammito, M., Usón, I., Read, R. J. & McCoy, A. J. (2018). Acta Cryst. D74, 245-255

Read, R. J. & McCoy, A. J. (2016). Acta Crystallogr. D. 72, 375–387.
Keywords: Likelihood, eLLG, LLGI