adding DFE for Arabidopsis

[chriscrsmith]

Addressing #1321

Using estimated gamma shape and mean from https://doi.org/10.1038/s41467-018-05281-7. This study estimates dominance in addition to the DFE, however I think we decided to roll with the additive-only model because it is easier to implement.

Note that I plugged in [0, 1.0] for the proportions of neutral and selected loci.

Also I put a minus sign on the gamma mean which seems to be convention.

[codecov]

Codecov Report

Merging #1324 (aac001f) into main (ec4fedd) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #1324   +/-   ##
=======================================
  Coverage   99.67%   99.67%           
=======================================
  Files         112      113    +1     
  Lines        3682     3697   +15     
  Branches      475      475           
=======================================
+ Hits         3670     3685   +15     
  Misses          8        8           
  Partials        4        4           

Impacted Files	Coverage Δ
stdpopsim/catalog/HomSap/dfes.py	100.00% <ø> (ø)
stdpopsim/catalog/AraTha/__init__.py	100.00% <100.00%> (ø)
stdpopsim/catalog/AraTha/dfes.py	100.00% <100.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[petrelharp]

One thing (and the reason codecov is complaining): you need also to put

from . import dfes  # noqa somethingsomething

in __init__.py.

[petrelharp]

And: this DFE is estimated without dominance? edit: Ah, I see from your comments that it is.

[petrelharp]

Note that I plugged in [0, 1.0] for the proportions of neutral and selected loci.

Is this right, or was it not clear from the paper?

[petrelharp]

Also I put a minus sign on the gamma mean which seems to be convention.

Yes, good - that is how you specify deleterious s in SLiM.

[chriscrsmith]

Note that I plugged in [0, 1.0] for the proportions of neutral and selected loci.

Is this right, or was it not clear from the paper?

I am pretty sure this is right. In the supp mat table with the gamma parameters I used, it does not list any proportion neutral. And the main text makes it seem to me that neutral mutations were not included in most DFEs estimates.

[chriscrsmith]

One thing (and the reason codecov is complaining): you need also to put

Doh. I must have gotten tired of waiting for the codecov to finish and I forgot to come back to this.

[chriscrsmith]

What does the comment "noqa ..." mean/do?

[grahamgower]

noqa

This tells flake8 to ignore the line. It can be used to ignore specific types of flake8 errors, e.g.

import foo  # noqa: F401

Means that flake8 won't produce the "imported but not used" error. The import may have desirable side effects, but flake8 doesn't know about this and just sees that the variable foo isn't used in the current scope.

[chriscrsmith]

noqa

This tells flake8 to ignore the line. It can be used to ignore specific types of flake8 errors, e.g.

import foo  # noqa: F401

Means that flake8 won't produce the "imported but not used" error. The import may have desirable side effects, but flake8 doesn't know about this and just sees that the variable foo isn't used in the current scope.

Got it! Thanks

[petrelharp]

Awesome, thanks! I"ll merge this; could you open a QC issue for it, @chriscrsmith?

[chriscrsmith]

QC issue #1344