25
edits
No edit summary |
(Reference to Ian Tickle et al. for sigma(Rfree)/Rfree = 1/sqrt(2n)) |
||
(3 intermediate revisions by 2 users not shown) | |||
Line 1: | Line 1: | ||
The following is based on a CCP4BB discussion around June 17, 2008 entitled: "How many reflections for | The following is based on a CCP4BB discussion around June 17, 2008 entitled: "How many reflections for [[iucr:Free_R_factor|R<sub>free</sub>]]" | ||
First of all, the test set is that set of reflections put aside for unbiased calculation of statistical quantities, in particular | First of all, the test set is that set of reflections put aside for unbiased calculation of statistical quantities, in particular [[iucr:Free_R_factor|R<sub>free</sub>]] and sigmaA. | ||
The need to find a good compromise for the size of the test set has been discussed by Axel Brunger in a "Methods in Enzymology" (1997) paper. He writes: | The need to find a good compromise for the size of the test set has been discussed by Axel Brunger in a "Methods in Enzymology" (1997) paper. He writes: | ||
Line 9: | Line 9: | ||
and the need to avoid a deleterious effect on the atomic model by omission of too much experimental data. | and the need to avoid a deleterious effect on the atomic model by omission of too much experimental data. | ||
==How precise is the estimate of | ==How precise is the estimate of R<sub>free</sub> for a certain number of test set reflections?== | ||
The estimate for the relative error of | The estimate for the relative error of [[iucr:Free_R_factor|R<sub>free</sub>]] is 1/sqrt(2n), where n is the size of the test set (Tickle et al., Acta Cryst. (2000) D56 , 442-450). So if n is 1000, and the [[iucr:Free_R_factor|R<sub>free</sub>]] is 31%, you would expect its relative error to be 31%/sqrt(2000), which is about 0.7%. | ||
==How many reflections do you need to get a good estimate of the sigmaA values (as a function of resolution) needed to calibrate the likelihood target?== | ==How many reflections do you need to get a good estimate of the sigmaA values (as a function of resolution) needed to calibrate the likelihood target?== |
edits