Let Philosopher print errors when there are protein hits from PSMs that cannot be found in the fasta file.

[fcyu]

This is a bug/feature I found by accident (https://github.com/Nesvilab/FragPipe/issues/537#issuecomment-995943414). Basically, when you search the data with FASTA file A, and run Philosopher to filter the result using FASTA file B, Philosopher won't print any warnings or errors although the protein entries are totally unmatched. The outcome is empty tsv files.

I think printing necessary error messages is good for a program to be robust and the result to be correct. That is how we find the protein assignment bugs in Philosopher based one the warnings from IonQuant (https://github.com/Nesvilab/FragPipe/issues/267, https://github.com/Nesvilab/FragPipe/issues/218): IonQuant performs sanity checks to ensure that the input from Philosopher makes sense.

Best,

Fengchao

[prvst]

We're aware of that, it's happening on purpose. Back a few years a go, we used to restrict the database to be the same as the one in the XML file headers. We changed that allowing any files because some people were running updated or modified versions of their database, without the need to run the entire pipeline again, and asked to let the program allow it, so we changed. We have to check with Alexey if he wants the previous behavior back into the program, or if we just add a warning.