Fix broken constructor chain after #1195 refactor — aim_ids, gff_gene_type and related params no longer forwarded to subclasses

[khushthecoder]

Description

PR #1195 (GH1194-species-config-refactor) refactored species-specific parameters to be read from the JSON config. However, as noted by @jonbrenas in their review comment, removing aim_ids, aim_palettes, gff_gene_type, gff_gene_name_attribute, gff_default_attributes, and gene_names from AnophelesDataResource.__init__() breaks the constructor chain — these parameters no longer reach AnophelesAimData or AnophelesGenomeFeaturesData.

---

Root Cause

The cooperative inheritance chain works like this:

Ag3.__init__() → AnophelesDataResource.__init__() → ... → AnophelesAimData.__init__()
                                                      → AnophelesGenomeFeaturesData.__init__()

Previously, ag3.py passed values like aim_ids=("gambcolu_vs_arab", "gamb_vs_colu") through AnophelesDataResource, which forwarded them via **kwargs to the subclasses.

After the refactor:

1. These parameters were removed from AnophelesDataResource.__init__(), so they're no longer forwarded
2. The JSON configs (v3-config.json, v1.0-config.json) do not contain AIM_IDS, AIM_PALETTES, GFF_GENE_TYPE, etc.
3. Both the config lookup and constructor fallback return None

---

Impact

Species	Parameter	Effect
ag3	aim_ids resolves to None	AIM calls (aim_variants(), aim_calls(), plot_aim_heatmap()) will fail
af1	gff_gene_type defaults to "gene" instead of "protein_coding_gene"	Genome feature queries return wrong results
af1	gff_gene_name_attribute defaults to "Name" instead of "Note"	Genome feature queries return wrong results
adar1, adir1, amin1	Same GFF params affected	Same incorrect defaults applied

---

Proposed Fix

Re-add the 6 removed parameters to AnophelesDataResource.__init__() and restore the corresponding values in the species files (ag3.py, af1.py, adar1.py, adir1.py, amin1.py). The subclass constructors (AnophelesAimData, AnophelesGenomeFeaturesData) already implement config-first-then-constructor-fallback logic, so this restores backward compatibility without undoing the config refactoring work.

Note: Moving AIM_PALETTES to the JSON config is not straightforward because the palette colors are computed from plotly at runtime.

---

Related

• Original issue: Refactor: Move hardcoded species-specific parameters to a unified configuration schema #1194
• Merged PR: Refactor: Move hardcoded species-specific parameters to SpeciesConfig #1195

[khushthecoder]

Hi @jonbrenas , I realized that I missed a few parameters in my previous PR(#1195), which affected the intended behavior. To address this properly, I’ve created a follow-up PR (#1210) that restores the correct parameter handling. Apologies for the oversight earlier, and thank you for your patience. Please let me know if any further changes are needed.

[khushthecoder]

Hi @jonbrenas, just a gentle follow-up on this PR when you get a chance.

This PR is a small, focused fix — just two lines:
Re-adds gene_names as an optional parameter in AnophelesDataResource.init()
Restores gene_names=GENE_NAMES in Ag3.init()
All CI checks are passing. Would you be able to take a quick look when you have a moment? Thank you for your time and guidance — I really appreciate it!