Dear Colleagues, I am preparing to deposit a cryo-EM structure to EMDB/PDB and have encountered an important question regarding resolution reporting and half-map deposition after Phenix density modification. My Situation: 1. I performed 3D refinement in CryoSPARC, generating two independent half-maps 2. I calculated FSC between these original half-maps at the 0.143 threshold: X Å resolution. 3. I processed these half-maps through Phenix resolve_cryo_em for density modification 4. Phenix reports improved resolution of Y Å (based on corrected FSC as documented in your resolve_cryo_em guide) 5. I generated the corresponding FSC XML file from Phenix showing Y Å Questions: When I deposit to EMDB, I understand the policy requires depositing the original, unmodified half-maps (unfiltered, unmasked, unsharpened). However, when EMDB validates the deposition by independently recalculating FSC from these half-maps, they will obtain X Å, not Y Å. This creates an apparent contradiction: • Deposited FSC XML file: Y Å (from Phenix) • EMDB independent FSC calculation from deposited half-maps: X Å (from originals) * When you apply corrections for correlations introduced by density modification in resolve_cryo_em, does this mean the Y Å resolution should be reported as the official resolution, even though the original half-maps show X Å? * What is the recommended workflow when depositing density-modified maps to ensure consistency and avoid validation conflicts? Additionally: • No atomic model was used; only sequence information • The density modification improved map quality and interpretability as intended Thank you. Kind regards, Dmitry
Hi Dmitry,
As far as I know, there is no official answer to your question.
Here is my logic and recommendation:
Logic:
1.Best to report analyses that are as concrete and reproducible as possible
and that refer to something that everyone can understand. Try to be
consistent with what has been done in the past. If the option is
available, try to add additional useful information.
2. In your case, the most widely-recognized resolution of your data is the
resolution of the original unmodified data (X).
3. Most would also agree that after density modification, the effective
resolution has been improved to Y.
Recommendation:
1. In cases where you only are allowed to report one number, report X. If
you can add information somewhere, report Y as estimated resolution after
density modification.
2. Submit your FSC curve from original unmodified data. If you have a way
to add information, submit the FSC values from ResolveCryoEM as well.
3. Note: do not calculate or report a standard FSC curve using
density-modified half maps. These maps have intrinsic correlation (removed
in the estimates made by ResolveCryoEM) so a value calculated directly from
them is not useful.
All the best,
Tom T
On Thu, Nov 20, 2025 at 4:17 AM Dmitry Semchonok
Dear Colleagues,
I am preparing to deposit a cryo-EM structure to EMDB/PDB and have encountered an important question regarding resolution reporting and half-map deposition after Phenix density modification.
My Situation: 1. I performed 3D refinement in CryoSPARC, generating two independent half-maps 2. I calculated FSC between these original half-maps at the 0.143 threshold: X Å resolution. 3. I processed these half-maps through Phenix resolve_cryo_em for density modification 4. Phenix reports improved resolution of Y Å (based on corrected FSC as documented in your resolve_cryo_em guide) 5. I generated the corresponding FSC XML file from Phenix showing Y Å
Questions: When I deposit to EMDB, I understand the policy requires depositing the original, unmodified half-maps (unfiltered, unmasked, unsharpened). However, when EMDB validates the deposition by independently recalculating FSC from these half-maps, they will obtain X Å, not Y Å.
This creates an apparent contradiction: • Deposited FSC XML file: Y Å (from Phenix) • EMDB independent FSC calculation from deposited half-maps: X Å (from originals)
* When you apply corrections for correlations introduced by density modification in resolve_cryo_em, does this mean the Y Å resolution should be reported as the official resolution, even though the original half-maps show X Å? * What is the recommended workflow when depositing density-modified maps to ensure consistency and avoid validation conflicts?
Additionally: • No atomic model was used; only sequence information • The density modification improved map quality and interpretability as intended
Thank you.
Kind regards, Dmitry _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
-- Thomas C Terwilliger Laboratory Fellow, Los Alamos National Laboratory Senior Scientist, New Mexico Consortium 100 Entrada Dr, Los Alamos, NM 87544 Email: [email protected] Tel: 505-431-0010
participants (2)
-
Dmitry Semchonok -
Tom Terwilliger