Update workflow for v1.0.0 release

[ergonyc]

NOTE: this is a post hoc excercise to demonstrate the process of creating release notes semi-automatically as the endpoint of development via github.

The previous minimum viable product version of the workflow employed R based tools and harmonized samples with the Harmony package. This workflow suffered from poor scaling properties: memory and resource utilization became un-manageable for greater than >500k cells.

Key updates:

• [ ] depricate R in place of python tools
• [ ] depricate snakemake handling
• [ ] rewrite scripts to leverage scanpy scvi-tools
• [ ] cellassign for infering cell types
• [ ] scvi for integration
• [ ] scib-metrics for tracking the tradeoff between technical variability due to batch/sample effects and conserving biological variability nescessary for further scientific inquiry

[ergonyc]

release

Cellranger v1.1.0 Cellbender + Counts to adata v1.0.0 Cohort analysis v2.1.0

Converted workflow from R to Python Used Cellbender to remove technical replicates Implemented scib for validation metrics