QC for AraTha DFE

[clararehmann]

Closes #1344.

I used the gamma distribution parameters from the additive A. lyrata model from Table S4 of Huber et al., 2018 to QC the model.

The scale parameter of the distribution is misidentified as gamma_mean in the catalog, so I modified the Gamma_H18 DFE in the AraTha catalog to match so that gamma_mean is now the product of the shape and scale parameters.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 99.84%. Comparing base (a344665) to head (33cbce7).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1647   +/-   ##
=======================================
  Coverage   99.84%   99.84%           
=======================================
  Files         133      133           
  Lines        4569     4580   +11     
  Branches      472      472           
=======================================
+ Hits         4562     4573   +11     
  Misses          3        3           
  Partials        4        4           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[petrelharp]

I've discussed this with @clararehmann; here's the summary.

As discussed in #1344, this DFE is arguably not the "best" one to use from this paper. However, it is a totally reasonable one, and who says we need just one DFE. So, the plan is to (1) add this additive DFE (and, this PR finalizes that); and (2) also add the "genome-wide both species h-s relationship" DFE (I will open another issue for that).

With that in mind: @clararehmann - how about let's change the name of this DFE, maybe to GammaAdditive_H18, as we discussed? (Or, make another proposal?)

[petrelharp]

Ah, uh-oh, @clararehmann - maybe there is a factor of 2 missing as well? Note that in that PR I agreed with you about the parameter being the scale, not the shape.

[petrelharp]

Ah - could you update the long description, also? How about

    Additive, deleterious DFE from Huber et al. (2018) for Arabidopsis, 
    estimated using <software> from A. lyrata as a Gamma distribution
    of deleterious effects. Parameters are from Supplementary Table 4,
    the "genome-wide, additive-only model for A. lyrata". The DFE for
    lyrata (rather than thaliana) is provided due to challenges with
    simulating selfing.

[petrelharp]

ah, I see <software> wasn't some named software package, they just did it using Methods from the paper. I rephrased this to "using the SFS"

[petrelharp]

Looks good! I've moved the improved description to the catalog, rather than the QC implementation (which no one will ever see); and squashed the commits. When the tests finish we can merge (assuming I didn't muck something up.)