Combine dataset
64 views
Skip to first unread message
Chen Sun
May 13, 2026, 6:24:20 PMMay 13
to Warp
Hello,
I have two small datasets that I processed with warp individually to tomogram alignment. How can I combine these two datasets for the following processes? Do I just copy the content in the warp_tiltseries into one folder?
Best wishes,
Chen
Reika Watanabe
May 13, 2026, 7:41:53 PMMay 13
to Chen Sun, 'Chen Sun' via Warp
Hi, Chen,
I am not sure that simple copy-and-paste will work.
Do the two datasets have the same pixel size? Do you plan to use these datasets for subtomogram averaging in Relion5 or others?
If they are the same pixel and camera dimensions, the simplest approach is to restart from preprocessing. That is what I usually do for all my datasets, so I do not have to think about it later. If you manually use Etomo for alignment, you can reuse that alignment.
If they have different pixel or camera dimensions (which is not the case for you), you may need to preprocess separately and combine the particles after particle extraction before subtomogram averaging.
I hope this helps!
Reika
--
You received this message because you are subscribed to the Google Groups "Warp" group.
To unsubscribe from this group and stop receiving emails from it, send an email to warp-em+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/warp-em/530445c4-b841-4f5b-bd68-9b96eb9d5971n%40googlegroups.com.
Chen Sun
May 13, 2026, 8:17:33 PMMay 13
to Reika Watanabe, 'Chen Sun' via Warp
Hi Reika,
Thank you for your reply. My two datasets were collected with the exact same pixel size. I guess it would be easier to merge the two dataset after picking the particles.
Best wishes,
Chen
Warp Bot
May 18, 2026, 1:15:30 AMMay 18
to Warp, Chen Sun
Hi Chen,
Adding to Reika's good advice: just copying the contents of warp_tiltseries alone won't work, because the XML files in there reference frame series XMLs (in warp_frameseries) and the underlying frames/gain. To merge at the WarpTools level you'd need to bring along everything: frames, mdoc, tomostar, warp_frameseries, and warp_tiltseries — plus a single settings file pointing at those folders. Since your tilt series names look unique across the two datasets (no filename clashes), this is technically fine, but only if both datasets share the same gain reference and the same pixel size.
For downstream work you actually have two clean options:
1. Restart from preprocessing as one combined project, as Reika suggested. This is the simplest if both datasets used the same gain reference.
2. Keep the two WarpTools projects separate all the way through subtomo export (ts_export_particles), then merge the resulting particle STAR files in RELION for refinement. When you move to M, add each dataset as its own data source in MTools — you do NOT need to physically combine the warp_tiltseries folders. M handles per-dataset CTF/gain/dose accounting correctly when datasets are kept separate. This is the recommended path if the two datasets have different gain references or were collected on different sessions.
For two small datasets with identical pixel size and (likely) the same gain reference, option 1 is the least error-prone. If gains differ, go with option 2.
— Warp Bot
THIS IS AN AUTOMATED MESSAGE GENERATED BY AN LLM. IT MAY OR MAY NOT SOLVE YOUR PROBLEM. IF YOU'D LIKE TO SPEAK TO A HUMAN, SAY SO IN YOUR MESSAGE.
Adding to Reika's good advice: just copying the contents of warp_tiltseries alone won't work, because the XML files in there reference frame series XMLs (in warp_frameseries) and the underlying frames/gain. To merge at the WarpTools level you'd need to bring along everything: frames, mdoc, tomostar, warp_frameseries, and warp_tiltseries — plus a single settings file pointing at those folders. Since your tilt series names look unique across the two datasets (no filename clashes), this is technically fine, but only if both datasets share the same gain reference and the same pixel size.
For downstream work you actually have two clean options:
1. Restart from preprocessing as one combined project, as Reika suggested. This is the simplest if both datasets used the same gain reference.
2. Keep the two WarpTools projects separate all the way through subtomo export (ts_export_particles), then merge the resulting particle STAR files in RELION for refinement. When you move to M, add each dataset as its own data source in MTools — you do NOT need to physically combine the warp_tiltseries folders. M handles per-dataset CTF/gain/dose accounting correctly when datasets are kept separate. This is the recommended path if the two datasets have different gain references or were collected on different sessions.
For two small datasets with identical pixel size and (likely) the same gain reference, option 1 is the least error-prone. If gains differ, go with option 2.
— Warp Bot
THIS IS AN AUTOMATED MESSAGE GENERATED BY AN LLM. IT MAY OR MAY NOT SOLVE YOUR PROBLEM. IF YOU'D LIKE TO SPEAK TO A HUMAN, SAY SO IN YOUR MESSAGE.
--
Reply all
Reply to author
Forward
0 new messages