Overlap in feature picking very low

[Simone Sidoli]

Dear everybody,

I have a problem in the feature picking when using SuperHirn.

Even though chromatography looks very reproducible and by using standard settings in "feature picking" I achieve very low overlap between runs. I can see it by downloading the aligned feature file in the "Alignment" page.

Only few features are detected in many runs of my experiment, also when I use wide retention time tolerance. Does anybody have some standard settings to suggest me?

Thank you very much.

[Micheleen Harris]

Hello Simone,

I usually use a  MS1 feature mz range max  in feature picking of 2000 and a min of 0.  How many features are you getting on average from the feature picking step?  Did you check the corra.log file for any errors?  Or the SuperHirn.log?  Let us know if those are showing some strange errors or results.

Best,

Micheleen

[Micheleen Harris]

Hello Again,

Two things:  Is your data from a profile mode or centroided mode on the mass spec?  Corra only works with profile mode data.  Also, there's a TPP tool called indexmzxml which will correct any mzXML indices which might be causing a problem.  It produces and output mzXML file with corrected indices.  This has worked for me in the past.

Best,

Micheleen

[Simone Sidoli]

Dear Micheleen,

thank you for your advices. I tried all of them, but unfortunately the results are still similar.

I am using profile Orbitrap data. Should I play maybe with the settings of peak score cutoff or intensity cutoff? I tried already LC score between 1 and 1000, but nothing has changed.

There is not much else I can think about.

Thank you again.

Simone