It is practically impossible. Even if you know molecular structure, these molecules may pack itself into crystal lattice depending on the temperature, crystallization conditions etc. Crystal structure prediction from the milecular structure isn't by any means straightforward.

Moreover, if the structure was solved 15+ years ago (before the area detectors and low temperature devices become commonly accessible) I strongly recommend remeasurement of the structure. You wrote it is a drug, so room temperature structure is important, but however, if you can use LT device, I suggest you to make LT measurement also and if the structure will be the same at both temperatures, take the better data.

hi, if you have the 2D molecular structure of the drug you can generate the 3D structure using for instance the prodrg server: http://davapc1.bioch.dundee.ac.uk/cgi-bin/prodrg

I am not so optimistic about it :) (but you made me curious, so I will give it a try)

Mario, I checked PRODRG but it is only for generating basic stereochemistry from 2D structure drawing. But Srinivas has 3D structure of *the molecule* yet, he asked about possibility to obtain *crystal structure* from it.

I did not understand that he looked for actual crystal coordinates... what would be the use of them? Are we talking of the coordinates of a molecule interacting with something or of the crystal of the molecule alone?

Once you have the coordinates you can plot XRD pattern and many other properties with software like VESTA. Bonds and angles are not an input in this kind of software.

What you need to do to get the coordinates in that case is DFT calculation.