xiVIEW not displaying top match for crosslink

[lostculture]

I am using this data set generated from a Mascot search: https://xiview.org/xi3/network.php?upload=11434-00297-47421-04190-96127 And then look at the crosslink between 336 and 419 with a score of 110, in the middle of the protein. This match is for a deamidated modification. There is another unmodified match with a score of 103 and 46 duplicate matches for the unmodified peptides. In the Selected Match Table the highest scoring match is 88. There are quite a few matches for this disulfide pair that have a higher score than the one reported as a top hit so it is not just an issue of not reporting the top hit. The second issue is that there are a lot less fragments identified in the xi spectrum view than in Mascot. This particular data set is ETD so you have to change the annotated peaks to ETD. All the missing fragment ions are identified as being doubly charged. The peaklist does not have a charge state for the fragment ions. Mascot Server automatically searches for matches to both singly and doubly charged fragments hence they are identified. I will try creating a deconvoluted peaklist as to see if that fixes this display problem.

[colin-combe]

Hi,

could you share the file you uploaded please? I'll check what's going wrong.

cheers, Colin

[lostculture]

Hi Colin,

Thanks for the quick reply. I actually work for Matrix Science in the US and have cc'ed by official email address.

You can download the files from here: mbf.co/biRMCk

It's a shortcut to MailBigFiles. Let me know if you need anything else.

Best regards, Richard Jacob

On 3/16/2022 1:04 PM, Colin wrote:

Hi,

could you share the file you uploaded please? I'll check what's going wrong.

cheers, Colin

— Reply to this email directly, view it on GitHub https://github.com/Rappsilber-Laboratory/xiView_container/issues/88#issuecomment-1069359825, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIINYATNGAVNICLKDU43GATVAIIABANCNFSM5Q4JRCJA. Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you authored the thread.Message ID: @.***>

-- Richard Jacob

"Science and everyday life cannot and should not be separated." Rosalind Franklin

[lostculture]

I tried searching with a deconvoluted peak list and the Mascot results are better but I still see the same issues in xiView.

[colin-combe]

Hi @lostculture

had a quick look at this.

Re. the first problem with the missing top match - i think the problem is really that the matches are not properly ordered by score? So the missing match with score 110.85 is there but its on the third page of the matches in the selected match table. I know there's a problem with the UI in that the gap at the bottom of that table can make it look like you're already seeing all the matches (when really there are more pages).

I'll fix the ordering of the matches by score in that table next week.

Re. the second problem with annotating the spectra - the CSV files xiVIEW reads don't actually contain all the information needed to reanntotate a spectrum. When you uploaded the CSV file you should (?) have seen a dialog asking which ions were searched for. Having the right options selected there should help with the ETD data.

A better option will be to upload an mzIdentML file rather than a CSV file. The mzIdentML file should contain all the info needed for correct annotations of spectra. However, a bug in xiVIEW means that isn't currently working correctly, it doesn't properly parse the ion types out of the mzIdentML file (sorry). We have a fix but it isn't yet deployed.

Last brief comment - we are already working with someone from MatrixScience in Europe ( @vrkosk ) to ensure that the mzIdentML files exported by Mascot are correctly read and displayed by xiVIEW. Hopefully, the reading of mzIdentML exported by Mascot should soon be quite reliable. I'll keep you in the loop about that.

Will be in touch about the ordering of matches by score.

Cheers, Colin

[colin-combe]

Hi Richard,

I think that's the ordering of matches by score working in that table,

cheers, Colin

[colin-combe]

~@lostculture Hi Richard, I found out last week that the "xiVIEW CSV export" from mascot doesn't reference the correct spectra for each identification. This may be why a lot less fragments are annotated in the spectrum viewer. I believe mascot are in the process of fixing this, best wishes, Colin~

[colin-combe]

hi again, what I wrote above was mistaken. The problem arose because the mgf had not been re-exported in query order. Its says to do this in the instructions at http://www.matrixscience.com/help/export_help.html#XIVIEW

col