Adds support for per-sample expected doublet rates in Scrublet

[AhmadAlalti]

Overview

• Introduces the feature to specify expected doublet rates on a per-sample basis for doublet detection using scrublet for heterogeneous datasets.
• Enables configuration through a CSV file mapping sample identifiers to their expected doublet rates.
• Maintains fallback to a global default rate if per-sample data is unavailable or missing for a sample.

Motivation

This update addresses the limitations of a single expected doublet rate by allowing tailored configurations, which can enhance doublet predictions when sample-specific rates are known or required.

Details

• Updates configuration options to allow toggling per-sample rate usage and specifying the CSV file path.
• Modifies pipeline logic to read sample-specific rates from the CSV file and revert to the default as necessary.
• Adds a resource CSV example for user reference.

[deevdevil88]

i have chatted with Ahmad, and he has tested extensively that this works