crystallography and biophysics, university of madras^* Lawrence Berkely laboratory, USA^**
○Gayathri Dasara Raju^* Zwart Peter^** Velmurugan Devadasan^*

The feasibility of SAD experiments can be assessed by the number of reflections with a significant Bijvoet difference and the substructure determination in any SAD experiment using modern direct methods program is directly related to the above number of reflections. The published work on the measurability of Bijvoet intensity ratio has a major drawback, namely, it lacks the incorporation of the effects of experimental errors. The strength of the anomalous signal can be judged by both the intensity and amplitude based anomalous signal to noise ratio and a quantity related to the average anomalous signal to noise ratio is the number or fraction of Bijvoet differences whose absolute value is larger than three times its standard deviation.

Recently, the relatively complicated integrations with the effects of errors on the expected Bijvoet ratio and measurability, which are not straightforward to solve by analytical means, have been bypassed by using a numerical approach. The numerical determination of the Bijvoet ratio and measurability in a SAD experiment involves the combination between the classical measurability and the anomalous signal to noise ratio criteria. The presentation will cover the estimation of the above in case of MAD data to assess the variation in measurability when one approaches the absorption edge of heavy atoms.