Fix import consistency and code structure in regression_2 consensus helper

[Copilot]

This PR addresses import inconsistencies and code structure issues in the src/metrics/regression_2/consensus/helper.py file to align it with the main regression_2 helper and repository coding standards.

Issues Fixed

Duplicate Import: The consensus helper was importing anndata twice:

import anndata          # Line 4
import anndata as ad    # Line 6 

Inconsistent Usage: Mixed usage of anndata and ad aliases throughout the code:

adata_rna = anndata.read_h5ad(par['evaluation_data'])  # Using full name
net = ad.read_h5ad(filepath)                           # Using alias

Import Organization: Imports weren't following Python best practices or matching the main helper's structure.

Changes Made

1. Removed duplicate import and standardized on anndata as ad (consistent with main helper and codebase)
2. Organized imports following PEP 8 guidelines:
   • Standard library imports first (os, json, sys)
   • Type imports (typing)
   • Third-party imports (numpy, pandas, anndata)
   • Local imports (util)
3. Added type hints to the main function for better code quality and consistency with the main helper
4. Improved code formatting with proper spacing and organization

Result

The consensus helper now has consistent import patterns with the main regression_2 helper while maintaining its specific functionality for generating consensus regulators data. All existing functionality is preserved - this is purely a code quality and consistency improvement.

Before:

import anndata
import anndata as ad
# ... inconsistent usage throughout

After:

import anndata as ad
# ... consistent usage throughout with proper organization

Original prompt

On helper.py, compare this to the main

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.