Understanding psm.tsv

[plbaldoni]

Hi there,

How exactly is 'Delta Mass' calculated in the psm.tsv output?

According to the Wiki, it is the difference between the calibrated observed peptide mass and the calculated peptide mass. I might have missed something in this calculation, since comparing 'Delta Mass' and the difference of the masses gives a remainder:

# In R:
> tb <- data.table::fread('../psm.tsv')
> summary(tb[,(`Calibrated Observed Mass` - `Calculated Peptide Mass`) - `Delta Mass`])
      Min.    1st Qu.     Median       Mean    3rd Qu.       Max. 
-0.0016000 -0.0011000 -0.0010000 -0.0009854 -0.0008000  0.0011000 

Also, why does psm.tsv contain several PSMs with PPM above the PPM that was specified during search (20 in the example below)?

# In R:
> tb[,sum(abs(1e6*(`Calibrated Observed Mass` - `Calculated Peptide Mass`)/`Calculated Peptide Mass`)>20)/.N]
[1] 0.1923914

Thanks for your support!
Pedro

[prvst]

Hi @plbaldoni. Philosopher don't actually calculate those values. These numbers come all the way down from the database search engine. Are you using MSFragger?

[anesvi]

Felipe, delta mass is calculated by philosopher. I think the discrepancy is because philosopher still does it’s own “zero bin” correction for the delta mass column (which, since msfragger now calibrates the data, not really necessary anymore)

…

Sent from my iPhone On Apr 15, 2021, at 8:25 PM, Felipe da Veiga Leprevost ***@***.***> wrote:  External Email - Use Caution Hi @plbaldoni<https://github.com/plbaldoni>. Philosopher don't actually calculate those values. These numbers come all the way down from the database search engine. Are you using MSFragger? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#202 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AIIMM6ZAGJI4E57JQV47EHTTI57YZANCNFSM4265LM2Q>. ********************************************************** Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues

[prvst]

We dropped the correction last year.

[plbaldoni]

@prvst Thanks. Yes, I am using MsFragger version 3.2. Would it help if I sent you the log and/or the params file?

[anesvi]

Hi Pedro, We are looking into this, but my understanding this is because both philosopher and msfragger are rounding off the values when printing them. We do not think there is any issue, but I cc Fengchao who can also clarify. Alexey From: Pedro Baldoni ***@***.***> Sent: Friday, April 16, 2021 10:04 AM To: Nesvilab/philosopher ***@***.***> Cc: Nesvizhskii, Alexey ***@***.***>; Comment ***@***.***> Subject: Re: [Nesvilab/philosopher] Understanding psm.tsv (#202) External Email - Use Caution @prvst<https://github.com/prvst> Thanks. Yes, I am using MsFragger version 3.2. Would it help if I sent you the log and/or the params file? — You are receiving this because you commented. Reply to this email directly, view it on GitHub<#202 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AIIMM6ZKFQYDYI7ZZY2EQLDTJA7WDANCNFSM4265LM2Q>. ********************************************************** Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues

[fcyu]

Hi @plbaldoni ,

As Alexey pointed out, the difference you observed regarding the Delta Mass is due to the rounding errors Philosopher only prints 4 decimal points. So the fourth digit of the values calculated with those columns are not 100% precise.

As to the PSMs with ppm outside of the tolerance, it is due to 1) the binary search used in locating the boundary; and 2) MSFragger used observed mass in converting ppm to Da internally. Here is an toy example of the issue in binary search.

Best,

Fengchao

[plbaldoni]

Thanks Fengchao!