v0.2.3

Limited lifelines to <0.30.1 to keep compatibility with python 3.10

v0.2.2

🆕 WHAT'S NEW IN v0.2

Major updates in this release:

• Cox regression support - Added survival analysis capabilities alongside logistic regression
• dsub integration - Built-in support for distributed computing on Google Cloud Platform
• Forest plot visualization - New main visualization option alongside Manhattan plots
• PEP-compliant naming - Changed to lowercase package/module names (affects import syntax)
• Expanded CLI support - Added command-line interfaces for cohort and phecode modules
• Simplified CLI commands - Added entry points for easier CLI usage (e.g., phetk phewas instead of python3 -m phetk.phewas)
• Default input/output file format - Changed from csv to tsv for better compatibility; csv is still supported but tsv is recommended
• Enhanced user experience - Various improvements for clarity and usability

v0.1.47

All of Us specific update:

• Updated .add_covariates() method in Cohort class to change naming convention of genetic principle components (pc) retrieved from All of Us pre-calculated PCA data: the first principle component will be labeled as pc1 instead of pc0 (previously used in v0.1.46 and earlier), and so on.

v0.1.46

• Updated components, e.g., Cohort module, to be compatible with the new All of Us data v8 release (Curated Data Repository (CDR) version 8 Release Notes).

v0.1.45

• Fixed a minor issue where Bonferroni line is shifted in Manhattan Plot after non-converged phecodes are removed.
• Updated README (Plot Module) to add description for other plotting features, such as setting custom Bonferroni value or custom color palette, etc.