[MASST] -Feature request - variable m/z range for analog search

[RaphaelR87]

Hi everyone,

as MASST is about to become the BLAST analog for mass-spec based metabolomics, I think the analog search is of utmost value. However, so far the analog search is limited to 100 Da max. While that allows for detecting many common biotransformation reactions of metabolites in the human body some very common biotransformations to 'detoxify' cannot not be detected such as conjugates of glucuronic acid and glutathion.

Furthermore, Natural Product Chemists would love to see analogs of their compounds of interest for discovery purposes but also to address envinronmental/ecological/evolutionary questions. The limit of 100 Da limits the usefulness of MASST for peptidic natural products dramatically (others of course too that have aglycon-glycoside pairs) as many amino acids are > 100 Da

Thanks for considering, it is a great tool already

Best, Raphael

[lfnothias]

Ming will probably provide a more comprehensive answer.

• The challenge is increasing to more 100 Da will be extremely slow and unsustainable.
• I would rather recommend to run molecular network with all the datasets that matched by MASST search. That feature is already available.

An other trick would be change the precursor ion mass value in search, but I don't recommend that ^^

[RaphaelR87]

Yeah, I was afraid that increasing computing time might have been the issue.

• ok so the re-analyze button will include all query-matching 'mzML files' and not only the 'specific scans of a file'? That is a great temporary solution for many cases (not in case a strain produces a very close analog of your compound, but not your original compound, which is quite commonly found)
• I still think a wider m/z range that can be pre-selected by the user would be very valuable. One of the most valuable properties of BLAST is that you can deduce phylogenetic relationships by comparing similar (not identical) proteins. And if a m/z range of 250Da would increase my analysis time from 30 min to let us a couple hours, it is totally worth it in my opinion
• thanks for the 'trick' I will give it a try

[lfnothias]

Do you know what modifications you are looking for ? or untargeted?

[RaphaelR87]

These are general suggestions. But in one case I am looking for analogs of a known and GNPS-annotated NRP. So modifications I am looking for amongst others are compounds with additional (non-canonical) AAs

[lfnothias]

So you expect multiple modifications with non-canonical AAs ?

[justinjjvanderhooft]

Some (most) non-canonical amino acids are > 100 Da....

On Wed, 22 Apr 2020, 21:36 lfnothias, notifications@github.com wrote:

So you expect multiple modifications with non-canonical AAs ?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/CCMS-UCSD/GNPS_Workflows/issues/496#issuecomment-617989968, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARV56AFXH67FO5XEFEL7ITRN5BMPANCNFSM4MOLBKSQ .

[RaphaelR87]

exactly, that was the point I wanted to make (thanks @justinjjvanderhooft). Actually I think that applies also to all canonical amino acids besides glycine, alanine, proline, serine. Following this thought in my opinion it is not just a nice feature to have, but more of a crucial parameter if you want to compare it to -and make MASST as successful as BLAST.

Do not get me wrong MASST is a fantastic idea and it is already great in its current status.

[lfnothias]

Thanks a lot for adding your input @justinjjvanderhooft

@RaphaelR87 , when you wrote "These are general suggestions." Did you actually meant "general modifications" !?

[lfnothias]

I think this is a very good case for the MASST prototype we discussed a while ago with @mwang87 and @XsirdanielX. @RaphaelR87 do you wanna help and test it by providing a list for the modifications you are interested in of: delta m/z expected / molecular formula for difference / name for these possible AA modifications ? Note that we already have simple general modifications (like oxidation etc), so no need to include them.

[RaphaelR87]

@lfnothias Haha sorry what I wanted to say was that my suggestions are of general nature, meaning everyone could benefit from a m/z range defined by the user no matter if the research question is used for targeted or untargeted metabolomics.

[RaphaelR87]

sure I can write a list, but that could become very extensive, I will write you an email

[justinjjvanderhooft]

Glycosylatios (162, 146, 132), glucuronidations (178), and some other interesting modifcations are all >100. Typically, when I use analogue search, I put it on 200….

Op 22 apr. 2020, om 22:47 heeft RaphaelR87 notifications@github.com het volgende geschreven:

@lfnothias https://github.com/lfnothias Haha sorry what I wanted to say was that my suggestions are of general nature, meaning everyone could benefit from a m/z range defined by the user no matter if the research question is used for targeted or untargeted metabolomics.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/CCMS-UCSD/GNPS_Workflows/issues/496#issuecomment-618031097, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARV56FGFQA6BR2TKZ3246DRN5JVZANCNFSM4MOLBKSQ.

[lfnothias]

@RaphaelR87 it is fine, you can provide a first non-extensive version just to test

[justinjjvanderhooft]

@RaphaelR87 - the table in the supporting information of this paper could be good starting point: https://www.nature.com/articles/nchembio.684 https://www.nature.com/articles/nchembio.684

Op 22 apr. 2020, om 22:51 heeft lfnothias notifications@github.com het volgende geschreven:

@RaphaelR87 https://github.com/RaphaelR87 it is fine, you can provide a first non-extensive version just to test

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/CCMS-UCSD/GNPS_Workflows/issues/496#issuecomment-618033389, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARV56HUBMX7XMCIWSE27ZDRN5KDPANCNFSM4MOLBKSQ.

[lfnothias]

Excellent @justinjjvanderhooft @RaphaelR87 no need for MF delta. Just m/z delta and name is fine.

[justinjjvanderhooft]

Whilst we are at it: https://pubmed.ncbi.nlm.nih.gov/24191063/ - also interesting mz deltas for saccharide moieties!

On Wed, 22 Apr 2020, 23:02 lfnothias, notifications@github.com wrote:

Excellent @justinjjvanderhooft https://github.com/justinjjvanderhooft @RaphaelR87 https://github.com/RaphaelR87 no need for MF delta. Just m/z delta and name is fine.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/CCMS-UCSD/GNPS_Workflows/issues/496#issuecomment-618038826, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARV56FDBJAMTCE7RR7QGRTRN5LPBANCNFSM4MOLBKSQ .

[RaphaelR87]

Great, yeah I was hoping that a list must already exists. Probably there will be an updated one from the Dereplicator paper and/or Norine database. I will check.

[RaphaelR87]

Amino_acids_delta_m_z_Name.txt

[RaphaelR87]

Norine DB has 543 monomer entries which are listed here: https://bioinfo.cristal.univ-lille.fr/norine/rest/monomers There probably is a way to get only the delta m/z and names out of these. https://bioinfo.cristal.univ-lille.fr/norine/rest/monomers/flat/xml https://bioinfo.cristal.univ-lille.fr/norine/rest/monomers/flat/json

[lfnothias]

Lets ask @alexeigurevich @hoseinmohimani