IIRC the previous discussion was about automated real-space refinement against maps calculated with or without the free set. I doubt if manual map-fitting is powerful enough to significantly bias the structure wrt Rfree set. To the extent that it is, refinement to convergence in reciprocal space with the free set excluded should remove the bias. And omitting the free reflections (= setting to zero) can be very bad for maps (unless the fillin option is used). Its a question of biasing the map toward what you hope is right vs biasing it towards what you know is wrong. eab On 04/21/2015 05:53 AM, Tristan Croll wrote:
Hi all,
In the phenix.refine GUI, the default options for the maps auto-generated in Coot include the R-free set in the map coefficients. Is this the best choice? It's fairly natural for the user to immediately inspect/correct their structure using the helpfully-provided Coot window - which, as per a recent discussion thread on this list, can easily lead to over-fitting. I suspect I've been looking at such a case over the last week or so - a mid-3A structure refined in Phenix with an amazingly low Rfree (0.234) and Rfree-Rwork (~0.016), despite clear (albeit non-disastrous) errors in a number of loops, and moderate-poor statistics in structural quality metrics. Tellingly, the Rfree goes up quite substantially upon re-refinement after correcting these issues. If maps generated with the Rfree set included should only ever be used for final display, wouldn't it be better to have them excluded by default?
Best regards,
Tristan
Tristan Croll Lecturer Faculty of Health School of Biomedical Sciences Institute of Health and Biomedical Engineering Queensland University of Technology 60 Musk Ave Kelvin Grove QLD 4059 Australia +61 7 3138 6443
This email and its attachments (if any) contain confidential information intended for use by the addressee and may be privileged. We do not waive any confidentiality, privilege or copyright associated with the email or the attachments. If you are not the intended addressee, you must not use, transmit, disclose or copy the email or any attachments. If you receive this email by mistake, please notify the sender immediately and delete the original email.
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]