Skip to content

Releases: DNAReplicationLab/pynanalogue

Release list

0.1.7

Choose a tag to compare

@sathish-t sathish-t released this 10 Feb 09:53

Changed

  • Removed redundant whitespace trimming for read_filter in Python bindings (handled by nanalogue library)
  • Updated README code examples to show complete, reproducible output matching actual program behaviour
  • Bumps nanalogue version to v0.1.7, this fixes a peek bug when zero len reads are encountered.

Added

  • Version check job in PyPI publish workflow (internal GitHub Actions change) to verify tag, Cargo.toml, Cargo.lock, and pyproject.toml versions match before building wheels
  • Automated README example testing (tests/test_readme_examples.py) that extracts code blocks and expected output from HTML comment markers in README.md and verifies they match

0.1.6

Choose a tag to compare

@sathish-t sathish-t released this 21 Jan 07:54

Added

  • New peek() function to extract BAM file metadata (contigs and modifications) without full processing
  • New seq_table() function to extract read sequences and base qualities for a genomic region as a Polars DataFrame, with insertions shown as lowercase, deletions as periods, and modifications as 'Z'
  • New tag parameter for polars_bam_mods, read_info, window_reads, and seq_table functions to filter by specific modification type (e.g., single-letter code "m" or ChEBI code "76792")
  • New win_op parameter for window_reads() function supporting "density" (default) and "grad_density" modes to measure modification density or gradient within each window
  • PyPI publishing via GitHub Actions using OIDC trusted publishing
  • Comprehensive filtering parameter tests for polars_bam_mods, read_info, window_reads, and seq_table functions
  • Gradient window tests with new example BAM files (example_10, example_11)
  • New two_mods_bam test fixture for testing multiple modification types
  • New pynanalogue-specific test data files for seq_table testing

Changed

  • Removed flaky 10% sampling test case in seq_table testing due to stochastic fluctuations with small sample sizes
  • Updated nanalogue dependency from 0.1.4 to 0.1.6
  • Consolidated build-wheels.yml into publish_to_pypi.yml workflow
  • Re-enabled musl benchmark job in CI
  • Renamed test_filtering.py to test_polars_bam_mods_filtering.py for clarity

Fixed

  • Fixed mismatch generation bug when simulating BAM files (via nanalogue 0.1.6)
  • Fixed full_region misinterpretation when an open-ended interval or entire contig was provided in the region parameter (via nanalogue 0.1.6)