This is the pepXML (not the mzXML) that is causing problems. From the name of the function in the stack trace:

PeptideprophetSummaryImpl$DistributionPointImpl.getModel3NegDistr()

It seems like PeptideProphet may be the source of the problem. The string "-1.#J" is a pretty odd value for a number.

How large is the .pep.xml file? If it is not huge, perhaps you could attach that here.

Okay,

Here is the pep.xml file. Thanks for looking at this.

Pretty odd. PeptideProphet clearly failed to model the 3-charge peptides. Not a problem, since it is a tiny distribution with a high point of 5. But, rather than just list all zeros for negative and positive modelled distribution, as it did for 1-charge peptides, it has 0 and -1.#J (positive and negative model respectively).

Seems like a bug in PeptideProphet. If you add the .xtan.xml file to this thread, I'll see whether the bug jumps out at me, or try to send it on to other TPP developers.

I should note that I have seen 3-charge distribution failures that did not cause this problem.

Thanks, here is the .xtan.xml