DIANN output with Triqler search engine score

[ypriverol]

Description of feature

@qinchunyuan I'm reviwing some of the DIANN Triqler outputs and I have observed that for some peptides the score is 0.0, Example:

proteins	peptide	charge	intensity	run	condition	searchScore
RL4_HUMAN	AAAAAAALQAK	2	21465.4	N294-2	lung	0.0
RL4_HUMAN	AAAAAAALQAK	2	31655.7	N295-1	lung	0.0

How are you getting those scores? Are those q-values? or probabilities?

[daichengxin]

Previously 1 - PEP. Now updated to searchScore= -np.log(Q.value)

[ypriverol]

For all DDA and DIA.? @timosachsenberg uses 1-qvalue when exporting triqler in PLFQ. Why we are doing log.

We should get there as much consistency as possible.

[daichengxin]

For DIA. I saw https://github.com/statisticalbiotechnology/dia_sum/blob/main/scripts/result_scripts/diann_output_to_triqler_output.py

[jpfeuffer]

No, we are using 1-PEP. Probably Diann does not have probabilities, so they needed to come up with a hack. We should follow their hack for now.
A triqler input of 0 (probability) means q-value of 1, which is fine. It won't really work for good scores, which can have scores higher than 1 now. But maybe triqler can handle this? Or they have another additional transformation after that?