Reanalysis of DIA data using DIA-generated library resulted in signigicantly less IDs.

[andyzim111]

We have acquired 18 phospho-samples on Thermo Orbitrap Fusion lumos using DIA. First, we analyzed 18 runs using the library-free strategy, which resulted in 26437 phospho-precursors. In this analysis, we generated a spectral library from these 18 runs. Then, we used this spectral library to reanalyze the same 18 runs with the same parameters as above, leading to 15697 phospho-precursors. What is going wrong with my analyses? log files have been attached.

20230117_SeV_CaTiO3_predicted.log.txt 20230117_SeV_CaTiO3_predicted_second.log.txt

[vdemichev]

The [20230117_SeV_CaTiO3_predicted.log.txt] is the correct analysis and yields the desired result - this is how it's supposed to be. DIA-NN generated a library and used it automatically to reanalyse the data (in the process of MBR). The [20230117_SeV_CaTiO3_predicted_second.log.txt] uses this highly project specific library for an analysis with MBR turned on, this is incorrect as MBR should not be used with highly specific libraries.

Best, Vadim