fix: add missing commas in Alliance column lists to prevent silent string concatenation

[AaryanCode69]

Summary

Fix missing commas in Alliance of Genome Resources column definition lists that caused Python's implicit string concatenation to silently merge adjacent column names into single malformed entries.

Problem

src/data_generation/alliance/__init__.py defines column name lists for parsing Alliance TSV files. Two lists had missing commas between adjacent string literals, triggering Python's implicit string concatenation — a language feature where "A" "B" silently becomes "AB" with no error.

Location 1 — molecular_interaction (line ~141)

# Before: 1 entry instead of 2
"Host organism(s)" "Interaction parameter(s)",
# Produced: "Host organism(s)Interaction parameter(s)"

Location 2 — genetic_interaction (lines ~180–184)

# Before: 5 strings silently concatenated into 1
"Annotation(s) interactor A"
"Annotation(s) interactor B"
"Interaction annotation(s)"
"Host organism(s)"
"Interaction parameter(s)"
"Creation date",
# Produced: one mega-string + "Creation date" → 2 entries instead of 6

This would cause incorrect column-to-name mapping when parsing Alliance TSV files, with every column after the concatenation point shifted by 1 (Location 1) or 4 (Location 2).

Fix

Added the missing commas after each string literal so they are treated as separate list entries:

# Location 1
- "Host organism(s)" "Interaction parameter(s)",
+ "Host organism(s)",
+ "Interaction parameter(s)",

# Location 2
- "Annotation(s) interactor A"
- "Annotation(s) interactor B"
- "Interaction annotation(s)"
- "Host organism(s)"
- "Interaction parameter(s)"
+ "Annotation(s) interactor A",
+ "Annotation(s) interactor B",
+ "Interaction annotation(s)",
+ "Host organism(s)",
+ "Interaction parameter(s)",
  "Creation date",

Changes

File	Change
src/data_generation/alliance/__init__.py	Added 6 missing commas across two column definition lists (molecular_interaction, genetic_interaction)

Impact Assessment

• No runtime breakage — the current upload_to_chromadb() loop only processes filetype == "genes", so these column lists are not yet consumed in production.
• No deployment risk — purely additive comma insertions; no logic changes, no new dependencies.
• Future-proofing — when the Alliance pipeline is expanded to process molecular_interaction and genetic_interaction data, these lists will now correctly match the actual TSV headers from the Alliance API.

Related Issue

Closes #118