Class FDR (Please do not merge, this yet)

[ypriverol]

This PR contains the FDR-class logic to validate novel peptides:

• [x] FDR-class implementation for idXML
• [x] Refactoring the command line config options.
• [x] Code cleaning.
• [x] 0.0.19 -> 0.0.20

[ypriverol]

• Number of PSM after Global FDR filtering: 7563
• Number of PSM after Non-bayesian FDR filtering: 7434
• Number of PSM after Bayesian FDR filtering: 7563

[timosachsenberg]

Exact same value for Bayesian? Ist that by design

[ypriverol]

To be honest, it shouldn't be, but I'm not expert in the Bayesian method. I will wait for the review of @yafeng @husensofteng for feedback.