From: Andrew Dalke (
Date: Mon May 07 2018 - 11:10:24 CDT

On May 7, 2018, at 14:33, Giacomo Fiorin <> wrote:
> PyMOL's original developer (tip: look his bio up) felt that there wasn't a good open-source 3D software for structural biology and drug discovery. Whether he thought that the VMD license was too restrictive or that it had different technical requirements, for sure the two codes have a good chunk of overlap but different focus. Note that VMD originally stands for "Visual Molecular Dynamics". If you do information-based modeling more often than physics-based you may find PyMOL extremely useful.

Yes, it was both of those and more. Remember, the VMD at the time was more limited than the VMD now.

VMD's graphics then - I'm thinking specifically of the ribbons, which I wrote - were functional, but clunky. PyMol's were better. VMD's graphics have become much better.

Warren came out of Axel Brunger's lab, so PyMol supported crystallography visualization well before VMD did.

(As an historical aside, the VMD selection language was influenced by Brunger's X-PLOR, as that was one of the MD programs we used before we wrote NAMD.)

People doing drug discovery wanted a way to visualize docked poses. VMD couldn't really handle that. I suspect PyMol could.

Warren wanted Python, plus a command language that could be mapped to Python. VMD at the time was only Tcl.

Warren also wanted the UI to be more integrated in the screen with the display, while VMD's form interface was based on NeXT-style inspectors.

I didn't use PyMol enough to remember the things that VMD could do that PyMol couldn't do.