Searching for a tool which calculates a MS/MS spectrum metric to tell how well it may be identified

[Dominik Lux]

Hey all,

Does someone know of a tool, which is able to measure the quality of single MS/MS spectra "how good they are identifiable"?

So a measure similar to the following: [0, 1] with 1 --> should be definitely identifiable and 0 --> the spectrum is just noise

[david.b...@ipbs.fr]

Hi Dominik,

I can think about two options but maybe more options exist (in OpenMS?):

- there is the QualScore tool which is part of the TPP distribution (http://tools.proteomecenter.org/software.php). But maybe it is a bit outdated now.

- there is this Recover tool (https://github.com/LSMBO/recover-fx), which is more recent I think and has a graphical interface.

I think both of them are open-source, QualScore I think is written in C++ and Recover in Java.

Example of filter implemented in Recover: https://github.com/LSMBO/recover-fx/blob/master/src/main/java/fr/lsmbo/msda/recover/gui/filters/LowIntensityThresholdFilter.java

They may not produce exactly the result you want but maybe you can adapt these tools to your own need.

And finally, I have never tested these tools but I know well Alexandre Burel who supervised the Recover developments.

He is on the EuBIC Slack if you have some related questions.

Best,

David