Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Evaluation of the First Batch of Elaborated Fragments from Cluster 6 and 15 for MurE #16

Open
LizbeK opened this issue Feb 7, 2020 · 11 comments

Comments

@LizbeK
Copy link
Contributor

LizbeK commented Feb 7, 2020

Update 07/02/2020:
I have manually looked through all the data sets focusing on the Cluster 6 and 15 pockets and couldn’t see anything bound in these sites.
We are currently dealing with some challenges with the data analysis of these specific datasets with the PanDDA software. Due to the low symmetry of the crystals (P1) the analysis takes very long and we get a lot of false results due to subtle rearrangements in certain side chains and loops which we don't understand have a good handle on. This is not the same conformational change observed between the apo and substrate bound forms. We are looking at clustering the data to see whether we can pin down some of these more subtle rearrangements.
The good news is PanDDA version 2 is scheduled for release soon, which we already know will speed up the analysis process a lot.

@mattodd
Copy link
Contributor

mattodd commented Feb 7, 2020

Hi @LizbeK . Thanks for the update. So am I right in thinking that despite the technical challenges you've had, you're nevertheless confident that none of the fragments we shipped have bound? i.e. we should consider them "inactives". It's not as though we ought to be waiting for further analysis and/or experiments?

@LizbeK
Copy link
Contributor Author

LizbeK commented Feb 7, 2020

@mattodd @danaklug @fidiris Consider them "inactives". Sorry.

@mattodd
Copy link
Contributor

mattodd commented Feb 10, 2020

OK, @fidiris and @danaklug are going to collate what we know and try to make sense of the results, which are quite surprising. One last query, relevant to the question we were discussing about whether there was non-overlapping substrate bound in the crystal. The assay is done in such a way that if substrate needs to be bound (i.e. ATP or amino acid or sugar) that it could be bound? i.e. it matters that we're manipulating protein in this assay that could have non-overlapping substrate if that is needed. Apologies if this is a naive question, I don't have details of the assay to hand.

@fidiris
Copy link
Contributor

fidiris commented Feb 11, 2020

Myself and @danaklug have now created a new page on the wiki to summarise the fragment analogues which were shipped for evaluation back in December. Hopefully having all the structures in one place will make it easier to understand the rationale behind why those compounds were synthesised.

The results for cluster 6 were particularly suprising for us. Having looked at the fragment analogues synthesised, a lot of them are structurally very similar to the original hits.

Would it be worth sending samples of the original hits for next rounds of screens as controls?

@drc007
Copy link
Contributor

drc007 commented Feb 11, 2020

Would it be worth sending samples of the original hits for next rounds of screens as controls?

I agree

@bendndi
Copy link

bendndi commented Feb 11, 2020

Would it be worth sending samples of the original hits for next rounds of screens as controls?

If you're also able to get a new sample of the original hits I'd suggest testing that too.

The page on the wiki looks good. You may have already done this but is there also a plan to have all this info collated in a single "live and updatable" file, ideally an sdf file or csv/ smiles file for use in datawarrior / excel ?

@LizbeK
Copy link
Contributor Author

LizbeK commented Feb 11, 2020

I would actually suggest "new / fresh" sample of the original hit on the same echo plate the compounds are going to be on; diluted and pipetted by the same person who prepares the rest of the samples. In that case it will be a good controle all the way through the experiment, which includes handling, shipping, storage, echo transfer, and the actual soaking.

@fidiris
Copy link
Contributor

fidiris commented Feb 11, 2020

If you're also able to get a new sample of the original hits I'd suggest testing that too.

Yes, so I've synthesised one of the hits for MurE (OSA_000005) and with the next MurD shipment I'm in the process of making or buying the original hits. We will do the standard characterisation (1H and LCMS) and check the purity before shipping for commercial compounds.

There is currently a google sheet that myself and @danaklug have been updating with our compounds. Is that what you mean? or do we need another file?

@LizbeK
Copy link
Contributor Author

LizbeK commented Feb 11, 2020

OK, @fidiris and @danaklug are going to collate what we know and try to make sense of the results, which are quite surprising. One last query, relevant to the question we were discussing about whether there was non-overlapping substrate bound in the crystal. The assay is done in such a way that if substrate needs to be bound (i.e. ATP or amino acid or sugar) that it could be bound? i.e. it matters that we're manipulating protein in this assay that could have non-overlapping substrate if that is needed. Apologies if this is a naive question, I don't have details of the assay to hand.

To answer @mattodd earlier question, for the XChem screen the crystals are prepared in the absence of any substrates and the soaking is also done in the absence of substrates.

@drc007
Copy link
Contributor

drc007 commented Feb 12, 2020 via email

@LizbeK
Copy link
Contributor Author

LizbeK commented Feb 12, 2020

Hi, Did you get crystals for all samples? Or did you get some compounds causing melting of crystals?

@drc007 I had no melting of crystals in this batch. I attach the relevant columns of the database file. The column you want to look at is MountingResult (column F). The comments looks something like this:
OK:Clear:Normal

First comment is either OK (mounted) or Fail (not mounted). All the soaked crystals were mounted.
Second comment describes what happened to the compound added to the drop. Clear (can't distinguish it from crystal drop), Phase separation (not mixing with crystal drop), Precipitated out, Crystalline (crystallized out), No comment (same as normal)
Third comment describes what happened to the crystal. Normal, Cracked, Melted, Jelly, No comment.

2020-02-12_Follow-up 1 soakDB file.xlsx

I noticed that 15 of the crystals didn't diffract (column G-I). I highlighted them in yellow.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

5 participants