add extra robust regression parameters

[tdelhomme]

We have an option extra_robust_gl, to make the regression only on samples possibly being errors (we remove samples with high allelic fraction and run the regression on remaining ones). This was implemented because true variants can biased the regression if present in a high proportion.

To better manage the extra robust regression, we need to control the minimum allelic fraction to exclude a sample at a position in the regression, and also the minimum and maximum proportion of samples having such an allelic fraction, to exclude them all. For that, 3 parameters have to be included:

• min_af_extra_robust
• min_prop_extra_robust
• max_prop_extra_robust

These parameters are already customizable in the regression here, but they need to be passed to the nextflow script here and here, and read by the R script here.

Obviously these changes have to be included in the doc!

[V-Catherine]

It's done :-)

[tdelhomme]

@V-Catherine : The best way is to mention the issue in the commit which fixed it (with #), and then ask for a pull request (to merge dev into v1.1b). This will closed the issue and report it in the log file !

[mfoll]

@aurelieGabriel could you please add the new parameters to the bash version? (when you come back). See #165