Incompatible Windows region configuration breaks FragPipe 21.0

[WeiqiangChen]

• Describe the issue or question:

When I ran LFQ-MBR workflow with slight modification in FragPipe 21, it showed the following error and exited before finishing.

Exception in thread "main" java.lang.NumberFormatException: For input string: "0,2876"
at java.base/jdk.internal.math.FloatingDecimal.readJavaFormatString(Unknown Source)
at java.base/jdk.internal.math.FloatingDecimal.parseFloat(Unknown Source)
at java.base/java.lang.Float.parseFloat(Unknown Source)
at com.dmtavt.fragpipe.tools.percolator.PercolatorOutputToPepXML.percolatorToPepXML(PercolatorOutputToPepXML.java:340)
at com.dmtavt.fragpipe.tools.percolator.PercolatorOutputToPepXML.main(PercolatorOutputToPepXML.java:59)
Process 'Percolator: Convert to pepxml' finished, exit code: 1
Process returned non-zero exit code, stopping
Cancelling 76 remaining tasks

The same files and settings worked fine with FragPipe 20.

• Upload your log file
  (If a log file hasn't been generated, go to the 'Run' tab in FragPipe, click 'Export Log', zip the resulting "log_[date_time].txt" file to avoid truncation, then attach the zipped file by drag & drop here.)
  log_2023-12-18_12-36-00_FragPipe20.zip
  log_2023-12-20_13-30-35_FragPipe21.zip

[fcyu]

I believe the error was due to two reasons:

1. Your computer has the configuration that uses , but not . as the decimal point.
2. The new version of Percolator decides to follow your computer's configuration, while the old version always used . as the decimal point.

Could you change the language and region configuration of your computer to make sure that it follows the common format, and try again?

Thanks,

Fengchao

[WeiqiangChen]

That is also my first impression. But I always set the decimal to "." not "," as this screenshot shows.

Maybe it uses my geolocation and sets it to "," ignoring the Windows settings.

[fcyu]

Can you change the format to English (United States), reboot your computer, and try again?

Best,

Fengchao

[WeiqiangChen]

Now it works after following your suggestions.

Thanks!
WeiQiang

[fcyu]

Thanks for your positive feedback. Then, I will close this ticket.

Best,

Fengchao

[guoci]

@fcyu I think this error originates from MSBooster. There is no Locale.setDefault(Locale.US); in MSBooster, so MSBooster uses the user's locale. See Nesvilab/MSBooster#4

[fcyu]

Then, why 1) version 20.0 does not have this issue? and 2) Percolator didn't crash?

Best,

Fengchao

[guoci]

From the log file, I can see the code is reading columns named either bray_curtis or unweighted_spectral_entropy, which is generated by MSBooster, not Percolator. https://github.com/Nesvilab/FragPipe/blame/develop/MSFragger-GUI/src/com/dmtavt/fragpipe/tools/percolator/PercolatorOutputToPepXML.java#L319-L340

[fcyu]

That's a good point! I also discovered that Percolator actually ignored those two scores generated by MSBooster

 Split1	 Split2	 Split3	FeatureName
 0.0000	 0.0000	 0.0000	rank
-1.0434	-0.9890	-0.9601	abs_ppm
-0.5347	-0.4921	-0.5887	isotope_errors
-1.1022	-1.2666	-1.2323	log10_evalue
 2.4467	 2.7693	 2.1834	hyperscore
 0.9478	 0.9381	 0.8122	delta_hyperscore
-1.4724	-1.8152	-1.4649	matched_ion_num
-0.3029	-0.2489	-0.2001	complementary_ions
 0.8369	 0.8655	 0.9377	ion_series
-0.5626	-0.5748	-0.5745	weighted_average_abs_fragment_ppm
-0.0206	-0.0032	-0.0253	length_7
 0.0359	 0.0078	 0.0164	length_8
-0.0196	-0.0007	 0.0033	length_9_30
 0.0312	-0.0106	 0.0122	length_31
 0.0000	 0.0000	 0.0000	ntt
-0.5792	-0.5953	-0.5584	nmc
 0.0000	 0.0000	 0.0000	charge_1
-0.0857	-0.0910	-0.1112	charge_2
 0.0443	 0.0551	 0.0618	charge_3
 0.0850	 0.0808	 0.1141	charge_4
 0.0744	 0.0480	 0.0445	charge_5
 0.0442	-0.0065	 0.0428	charge_6
 0.0000	 0.0000	 0.0000	charge_7_or_more
 0.0000	 0.0000	 0.0000	group_1
 0.0000	 0.0000	 0.0000	group_2
 0.0000	 0.0000	 0.0000	group_3
 0.0000	 0.0000	 0.0000	group_other
-0.1570	-0.1799	-0.1676	15.9949M
 0.6170	 0.7373	 0.5897	m0

Best,

Fengchao