Hi everyone. We have just released SLiM 3.3.2. This is a minor release, with a couple of small additions:
- some performance improvements for large models and some other specific cases
- new recipes: 5.2.4 (joining subpopulations), 16.17 (meiotic drive), and 13.6 (a variety of fitness functions)
- revisions to existing recipes: 5.4 (the Gravel model) to show how to output a sample taken across multiple subpopulations, and 13.4 (a quantitative genetics model with heritability) to fix a bug introduced in SLiM 3.3 that broke the environmental noise contribution to phenotype
- other minor tweaks (see the release notes in the manual for more details)
One significant bug was found and fixed:
- if a non-singleton vector of positions was passed to AddNewDrawnMutation()/AddNewMutation(), and those positions were not in sorted order, the underlying data structures of the genomes for the target individuals would be corrupted, producing incorrect model results, because sorted order was assumed by the code; if this bug could apply to you, you should redo all of your runs, even if you didn't notice any issue. However, this should apply to very few end users; you would have to be explicitly calling AddNewDrawnMutation() or AddNewMutation() and passing a non-singleton vector of positions that is not in sorted order.
The other nice thing included in this release is necessary changes for SLiM to build under conda-forge. Rolling those changes into an official release should now allow us to make SLiM available through conda, for those who use that package installation mechanism; we will make a separate announcement once that is working. Building from sources for installing with the macOS double-click installer will continue to work as usual.
This upgrade is recommended for all users, since it fixes a serious bug and should be quite stable. This release preserves backward reproducibility (i.e., the same model, with the same seed, will produce the same results as SLiM 3.3.1, apart from the bugs fixed) and backward compatibility (models that ran under SLiM 3.3.1 should not require changes to their code).
You can obtain SLiM 3.3.2 from the SLiM home page at http://messerlab.org/slim/
; note that the manuals, recipes, and reference sheets have also undergone revisions.
If you have any questions, comments, etc., please use the slim-discuss group for that. Thanks, and happy modeling!
Benjamin C. Haller