Reply-to: Help and discussion for users of IMP <IMP-USERS@LISTSRV.UCSF.EDU>
If you are concerned about the GMM reproducing features of the original density, increasing the number of Gaussians is the best path forward. We don't usually adjust the other parameters. Note though that that will slow down the scoring since it scales linearly with the number of Gaussians.
Ben
--
https://salilab.org/~ben/
"It is a capital mistake to theorize before one has data."
- Sir Arthur Conan Doyle
________________________________________
From: Help and discussion for users of IMP <IMP-USERS@LISTSRV.UCSF.EDU> on behalf of Cecilia Sanders <>
Sent: Monday, October 26, 2026 2:36 PM
To: IMP-USERS@LISTSRV.UCSF.EDU
Subject: [IMP-users] Question about improving GMM.txt file quality
Hello, I am working on an integrated model utilizing a ~3. 4-angstrom cryo-EM structure and a crosslink dataset. Comparing the structure produced by my 400-gaussian gmm. txt file to my cryo map mrc, I am concerned that many details are lost in
ZjQcmQRYFpfptBannerStart
This Message Is From an Untrusted Sender
You have not previously corresponded with this sender.
ZjQcmQRYFpfptBannerEnd
Hello,
I am working on an integrated model utilizing a ~3.4-angstrom cryo-EM structure and a crosslink dataset. Comparing the structure produced by my 400-gaussian gmm.txt file to my cryo map mrc, I am concerned that many details are lost in translation. Are there ways to improve the precision of the gmm.txt file?
For instance, would any of the following tools from the usage guide be helpful, and how would I use them? [-t {spherical,tied,diag,full}] [-m OUT_MAP] [-a APIX] [-n NUM_SAMPLES] [-i NUM_ITER] [-s THRESHOLD] [-f FORCE_RADII] [-w FORCE_WEIGHT] [-e] [-d] [-k]
[-x CHAIN] [-z]
This is the command I am currently using in IMP 2.24:
python create_gmm.py \
cluster01.mrc \
400 \
cluster01_400.gmm.txt
Pictures left to right of my spherical, 1-MDa complex in mrc and gmm.mrc:
[cid:de9e8f63-b005-407d-ba32-14d0d74d903b][cid:3d1bb371-3486-4985-917b-39f701e0c4e6]
I did consult AI and received this recommended improvement. Would you add any suggestions to this?
python create_gmm.py \
-s <chimerax_viewing_threshold> \
-n 4000000 \
-i 300 \
-t full \
-m cluster01_400.gmm.mrc \
-a <map_resolution_in_angstroms> \
cluster01.mrc 400 cluster01_400.gmm.txt
Thank you so much for your help,
Cecilia Sanders
6th year Ph.D. student, Barry Willardson Lab
BYU Dept. of Chemistry and Biochemistry
[cid:496fdbb2-6b32-4a1b-97cd-0779abfbfee2]
________________________________
IMP-users<https://integrativemodeling.org/imp-users.html> mailing list; unsubscribe<mailto:imp-users-signoff-request@listsrv.ucsf.edu>
_______________________________________________
IMP-users mailing list
To unsubscribe send an email to imp-users-signoff-request@listsrv.ucsf.edu