Rotating stereo formulas with wedge bonds

Hi, thank you for your posts.


It is really a good forum to discuss problems. And the software to post a subject is well programmed. To have bold, color and all the other options is a nice feature. To add attachement in a comfortable procedure is fine. And the best feature is the "preview" option.





But know I have a new problem:


Inserting the template 2S-alanine to the MarvinSketch screen, and then using View-Transform-Rotate in 3D and rotating the formula around a horizontal axis produces formula 1, which is wrong. It is now the 2R-configuration. Rotating around the horizontal axis should produce a formula where the solid wedge changes to a hashed wedge. It is clear, it's a planar formula on the "paper plane", but the result is wrong and confusing.


And there is a second problem. Dragging now a second alanine from the Template Library gives a formula attached to the mouse pointer which is the 2R-configuration (see formula 2). And putting it to the drawing plane produces formula 3. What's wrong ?


To say two sentences more. My intention is not to get a workaround. The really good Marvin software should be usable in teaching courses.





I think this discussion should be a new topic. Should I post it in a new topic called "Rotating stereo formulas with wedge bonds" ?





(The screenshot is a real digital foto because I could not move the mouse pointer to make a software screenshot ;-)

Dear Hans-Ulrich,





I think Edit > Transform > Horizontal Flip rotates the structure correctly.


Horizontal Flip: Flips the selected object(s) horizontally, preserving the configuration of all enantiomers.





Regarding the problems you write, I think my colleague Akos will discuss this with you, though he is on holiday this week.





Kind regards,


Judit

Dear Judit,


Super ! That's what I wanted to have and to show.


Please look at the screenshot.





Thank you. Kind regards, Hans-Ulrich

Hi,





I think that 3D rotation is misleading for a structure with wedges, and it should be avoided. The wedge representation is only relevant as long as the diagram remains in 2D.





I suggest a 3D clean before rotating in 3D. (Menu item "Structure -> Clean 3D -> Clean in 3D" or a "Ctrl-3" keyboard shortcut.) Or simply use the flip functionality mentioned by Judit before.





Szabolcs

Hi Szabolcs, I must answer to your sentences:

Quote:

I think that 3D rotation is misleading for a structure with wedges, and it should be avoided. The wedge representation is only relevant as long as the diagram remains in 2D.

I agree to your second sentence. But concerning your first sentence I think 3D rotation is misleading for all formulas in 2D representation. So 3D rotation should be blocked for formulas in 2D , or the Clean 3D method should be done automatically before 3D rotation.


There are other features, where the Clean 3D method or the Prehydrogenisation method is done automatically before the calculation is done, for example in the "Geometry calculation".





Hans-Ulrich

Hi Hans-Ulrich,





First of all, let me state that I see your point: the 3D rotations of wedge bonds are presently problematic in Marvin.





Let me call your attention to the difference between the two 3D rotation feature presently available in Marvin:


- The 'Edit > Transform > Rotate in 3D' feature really rotates the selection in 3D (you can check in the source that the Z coordinates are different form zero after this action). After using this feature the sketch mode automatically switches to 3D mode.


- The 'View > Transform > Rotate in 3D' feature rotates only the view (and does not effect the coordinates). The sketch mode remains in 2D.





These features are active even in case of 2D structures, because they allow to create 3D projected drawings, even though these are not real projections yet (we are planning to implement bold bonds for projected views in the future). In addition, we also want to enable 3D sketching without initial 3D cleaning (someone might not want to clean his structure).





Unfortunately, both rotation types might result in undesired structure representation at the end:


- Since the "real rotation" does not convert the wedge bonds to single bonds, even though the S/R label disappears, the resulted sketch may be misleading. In addition, if I make a single 2D clean after, the stereo info is actually lost. (On the other hand, if I make first a 3D clean, then a 2D clean, the stereo info is retained to the original one). An immediate 3D clean would help, but it would demolish any projected drawing immediately. We will think about the solution.


- There is a bigger problem with the rotation of the view, because - as you pointed out - it does not change the wedge bond when rotating the view over the edge of the plane. It is a bug, and will be fixed soon.





Thank you for your notes again.





Best regards,





Akos