search

vdemichev / DiaNN

DIA-NN - a universal automated software suite for DIA proteomics data analysis.
Other
237 stars 50 forks source link

Questions About Some Columns #1068

Open Tan-Jiahua opened 1 week ago

Tan-Jiahua commented 1 week ago

Hi,

Thank you for developing DIA-NN! I have used DIA-NN (version 1.8.2) within the FragPipe platform (version 21.1). I have started exploring the main output and noticed that certain columns (Evidence, Spectrum.Similarity, Averagine, Mass.Evidence, CScore, Decoy.Evidence, Decoy.Score, Lib.Index, iIM, Predicted.IM, Predicted iIM, Fragment.Correlations) aren't described in the main output reference on https://github.com/vdemichev/DiaNN. While some of these columns can be inferred or discussed in other DIA-NN issues, I'm curious if their introduction is available.

Thank you for your time and assistance.

Best wishes

vdemichev commented 1 week ago

iIM refers to the reference ion mobility in the spectral library, IM to the empirically measured, Predicted.IM - IM predicted based on iIM, Predicted.iIM - iIM predicted based on IM. Averagine is the measure of how well the MS1 isotopic profile relfects the predicted one based on the averagine distribution. Spectrum.Similarity is a complex measure of how similar the spectrum is to the reference spectrum in the library. Fragment.Correlations are the correlation scores for individual fragments, you can obtain the fragment annotation using --report-lib-info. Other scores are not meant to be used.

Tan-Jiahua commented 6 days ago

Thank you very much for your answer and sorry for the late reply. I would have a look on them and I may ask if I do not understand some points