[chimerax-users] Manual segmentation in ChimeraX

[Tom Goddard]

Hi Pattana,

  That brain segmentation was not done in ChimeraX.  It used custom machine learning plus 4000 hours of manual corrections by students in custom editing software.

  But for much simpler cases a few ChimeraX approaches could do.  One is to place markers interactively with the marker mouse modes and then take all density within some distance with command volume zone.  Can also take density near an approximately fit atomic model, but that probably won’t be useful for lower res EM.  Another method is use the Volume Eraser tool to erase everything outside the region of interest.  It lets you move a sphere around and erase with it.

  These ChimeraX methods are really simplistic and usually not sufficient.  You would probably do better segmenting manually in IMOD.

   Tom

> On May 13, 2020, at 12:38 PM, Pattana Jaroenlak (Michael) <p.jaroenlak at gmail.com> wrote:
> 
> 
> Hello,
> 
> I saw the twitter post from ChimeraX on segmentation of EM serial images (https://twitter.com/ucsfchimerax/status/1260425173238415360?s=12). It was very beautiful. I have serial section EM images. I am wondering that can I do manual segmentation on ChimeraX? Some cellular features on my images can not be picked up by thresholding or watersheding.
> 
> Thank you
> PJ
> 
> -- 
> Pattana (Michael) Jaroenlak, PhD
> Bhabha & Ekiert Labs
> Skirball Institute of Biomolecular Medicine
> New York University School of Medicine
> Email: pattana.jaroenlak at nyulangone.org, p.jaroenlak at gmail.com
> _______________________________________________
> ChimeraX-users mailing list
> ChimeraX-users at cgl.ucsf.edu
> Manage subscription:
> https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimerax-users/attachments/20200513/76632dbe/attachment-0001.html>