jaxqtl 0.3.1

v0.3.1

on Python PyPI

Expression PCA

• Expand compute-pcs with sample and gene selection, expression-prevalence filters, optional covariate alignment, and external library sizes.
• Separate normalization (library-size, tmm, or none) from transformation (log1p or none). Defaults are library-size normalization followed by log1p.
• Add TMM composition adjustment, estimated before PCA gene selection, with tests against edgeR reference cases.
• Number components from ExprPC1 and order them by explained variance.
• Return explained-variance proportions from the Python API and save them to a companion .variance.tsv file for scree plots.
• Improve input validation and preserve sample order during filtering.

Documentation

• Split the CLI reference into overview, mapping, and expression PCA pages.
• Rewrite the PCA guide around worked examples, including scree plots from saved variance tables.
• Update the documentation logo and favicon.

Compatibility

The PCA interface now uses separate normalization and transform arguments. The combined lognorm option is removed; use --normalization library-size --transform log1p. To skip both stages, explicitly select --normalization none --transform none.

Install

pip install --upgrade jaxqtl==0.3.1