about SASBDB

Suggestions for new topics to discuss, questions on the use of the Forum
Post Reply
Message
Author
ddurand
Member
Posts: 6
Joined: 2012.02.14 17:35
Location: CNRS France

about SASBDB

#1 Post by ddurand » 2014.12.11 10:41

I’m having a problem by submitting data in SASBDB. The field “Molecular weight estimated from Guinier I(0)” is required. In my case the molecular weight cannot be estimated from Guinier I(0) because the macromolecule is a complex DNA-protein and the protein does not contain any tryptophan. Furthermore the experiments were performed using SAXS coupled to HPLC. Consequently the concentration is not well known. Anyway it is difficult to deduce the weight from I(0) for a complex DNA-protein.
I think that the field “Molecular weight estimated from Guinier I(0)” should be optional.
Thank you in advance for your answer.
Dominique

evalenti
Guest
Posts: 2
Joined: 2011.09.25 17:11
Location: EW

Re: about SASBDB

#2 Post by evalenti » 2014.12.11 13:07

Dear Dominique,
I understand your point but we need to have at least one molecular weight calculated from the SAXS experiment in order to asses the quality of the experiment.
Usually the MW based on the I(0) is the most common and this is why this field is mandatory.
Do you have the MW based on the Porod volume? You can put it in the "MW based on I(0)" field and specify what you did in the "short description" field.
Thanks a lot for asking,
Erica

Annette
Active member
Posts: 17
Joined: 2008.10.29 13:50
Location: University of Copenhagen

Re: about SASBDB

#3 Post by Annette » 2014.12.17 16:41

Deal all,

I have a question on how to submit ensemble models to the data base.

In the submission form for SASBDB, one can upload two fits, each with two models.
I would like to submit data with the corresponding ensemble models, which in our case includes up to 10 pdb files in the ensembles.
Although I see examples with multiple models in the data bank (e.g. SASDAR3) I fail to find the option to include more models.

How do I / Can I include multiple pdbs?
And if so is it possible to also include the frequency/weight of the given model?

best wishes,
Annette

PS You may consider a new forum topic on the SASBDB. It didn't really make sense to post it under the Feedback, but at least I could continue in a SASBDB related thread.

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: about SASBDB

#4 Post by Alex » 2014.12.17 16:55

Anette, i think for ensemble peDB is better - http://pedb.vib.be/
I have a good experience with submission, etc..
Cheers, Alex

Annette
Active member
Posts: 17
Joined: 2008.10.29 13:50
Location: University of Copenhagen

Re: about SASBDB

#5 Post by Annette » 2014.12.17 22:05

Hi Alex,

Thanks for tip. For some of our projects the PEDB may be better, but in this case I think not since it is neither IDP nor denatured proteins. In this case it is ensembles to describe the flexibility of antibodies, thus I think SASBDB may be more suitable in this case.

cheers,
Annette

ddurand
Member
Posts: 6
Joined: 2012.02.14 17:35
Location: CNRS France

Re: about SASBDB

#6 Post by ddurand » 2014.12.18 11:59

Hi,
I agree with Annette. I think that even in the case of rigid macromolecules it could be useful to put on SASBDB the ensemble of conformations of the lowest-chi structures obtained using BUNCH or SASREF for example.
Best wishes,
Dominique

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: about SASBDB

#7 Post by Alex » 2014.12.18 12:58

i saw the paper about antibodies with in-depth ensemble analysis. I don't really appreciate the results of dKratky when it is shown without
comparison to unfolded and folded proteins (fig 1C). However, given ensemble distributions shown are quite broad, it is clear that molecules are quite flexible (especially G1 and G4). It is your call whether to call them IDPs or not but if your modeling is relying on ensembles analysis and you wanna deposit ensembles, I think "Protein Ensemble Database" sounds for me like a right place for it.

On the other note - I noticed that EOM fits after ~0.25 are not that good. Do you have an explanation for that? I believe that it might be to some additional internal dynamics. Actually, i posted earlier on this forum that I am looking for examples like that. Basically, i might be able to improve the fits - let me know if you want to give it a try.

Best, Alex

Post Reply