Form-factor of different samples - to distinguish

[comcon1]

10.1103.008
there two form factors WITH ABSOLUTELY IDENTICAL README.yaml
https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.100.198103

but the SAMPLES ARE VERY DIFFERENT.

This question is likely related to this PR: NMRLipids/FAIRMD_lipids#352

[comcon1]

You can also see the consequences in this PR: #102
It is uncertain which FF to select.

[korbinib]

I think this could be solved by an additional metadata attribute on the sample type.
Ideally with a controlled vocabulary - but I am unsure whether is sufficiently covered in any existing ontology.

[comcon1]

Brief search showef that there is no controlled vocabulary for this. Should invent our own terms... Best regards, Alex

…

-------- Оригинальное Сообщение -------- 09.10.2025 09:50, Korbinian Bösl пишет:

korbinib left a comment [(NMRLipids/BilayerData#104)](#104 (comment)) I think this could be solved by an additional metadata attribute on the sample type. Ideally with a controlled vocabulary - but I am unsure whether is sufficiently covered in any existing ontology. — Reply to this email directly, [view it on GitHub](#104 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AGBXJOFCDCWI5Z5OS2EQTIL3WYHUPAVCNFSM6AAAAACIUK5UA6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGOBUGU4DKMZUGQ). You are receiving this because you authored the thread.Message ID: ***@***.***>

[korbinib]

Could not find any thin along these lines either.

Could we try to come up with

1. a list of experiment/measurement types (maybe there will be a match in CHMO)
2. A list of sample types that are usually measured here

[hsantila]

I somehow managed to comment the related pull request PR #102. Here is my comment again:

I'm confused cause there are no form factors in the paper you linked. I'd be interested to see the difference between measurements of identical composition & temperature but different sample type. That would help estimate how big of an issue this is.

In the general case there are multiple experiments per simulation, we either have to pick one or use all.

Case "pick one": there needs to be a strategy for picking. I'd suggest either choosing one with lowest error estimate or adding a field for sample description based of which we can prefer one over another. This would need more vocabulary and consulting experimentalists on which types of samples are used and which are more accurate. It is not clear to me which case is most identical to our simulated curves.

Case "use all": Plot all experiments in the gui and use averages for quality estimation. I'd still ad a field for sample description.

[comcon1]

'm confused cause there are no form factors in the paper you linked. I'd be interested to see the difference between measurements of identical composition & temperature but different sample type. That would help estimate how big of an issue this is.

https://github.com/NMRLipids/BilayerData/tree/main/experiments/FormFactors/10.1103.008/PhysRevE.80.021931

[comcon1]

Case "pick one": there needs to be a strategy for picking. I'd suggest either choosing one with lowest error estimate or adding a field for sample description based of which we can prefer one over another. This would need more vocabulary and consulting experimentalists on which types of samples are used and which are more accurate. It is not clear to me which case is most identical to our simulated curves.

Case "use all": Plot all experiments in the gui and use averages for quality estimation. I'd still ad a field for sample description.

liposomes has probably some shape factor and form factor and lamellar phase has probably only form factor. I'm not very sure whether there is a difference in processing that could affect the comparison. For the lamellar phase, you also have a "reflection" from the general periodicity. And from liposomes you don't. If they are ULV. If they are MLV, you probably have something.

[hsantila]

I think, if we are lucky, we may face a situation where there are several experiments of same/similar systems in the databank so adding more fields to differentiate systems and the ability to link simulation to more than one experiment might make sense in the future