Skip to content

v1.1.0
Compare
Choose a tag to compare
@AlexTate AlexTate released this 22 Oct 20:51
· 470 commits to master since this release
v1.1.0
1588297
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.

This release brings improvements in performance, reliability, and compatibility.

What's Changed

❗ = changes that may present issues with backward compatibility

Pipeline changes

  • All Conda dependencies have been updated, including Python (3.7 to 3.9), bowtie, fastp, matplotlib, etc. R dependencies are now managed by Conda which has removed lengthy build steps from the installation script. #214
  • GFF annotations are validated at both pipeline and tiny-count startup. These changes also bring expanded support for feature ID attributes (by priority: ID, gene_id, and Parent) and annotations defining whole chromosomes for compatibility with Ensembl. #236
  • Samples Sheet contents are validated at pipeline startup #243
  • ❗ Large bowtie indexes (*.ebwtl) are now supported. The activation steps for bowtie-build have also been simplified. #238

tiny-count changes:

  • A custom Cython implementation of HTSeq's StepVector has been introduced. As a result, tiny-count runtimes are up to 50% faster (depending on configuration). HTSeq's StepVector is still used as a fallback if there are issues during build or import, or via user preference in the Run Config. Note: prebuilt binaries are not provided in this release. #218
  • The semantics of the Features Sheet Hierarchy value have changed. Instead of being used as a means of candidate elimination in Stage 2 selection, it is now used to sort Stage 2 matches as a priority order for Stage 3. #229
  • Features are no longer required to be stranded. Unstranded features that match rules with 3'/5' anchored overlap selectors will be downgraded to a new anchored overlap selector which does not distinguish between 3'/5' ends. Unstranded features will match all strand selectors (sense, antisense, and both) #236
  • Features that list more than one value for their ID attribute are now accepted. These values are concatenated to form a single ID string. #239
  • Non-collapsed SAM files, as well as SAM files produced from fastx collapsed outputs, are now accepted by tiny-count. #217
  • ❗ The required command line argument for the Features Sheet has been renamed. #217
  • GFF parsing performance has been improved. #218
  • Gzipped tiny-collapse outputs no longer lead to a crash during Summary Stats creation #222

tiny-deseq.r

  • Excessive decimal precision has been fixed in the pvalue and padj columns of DGE tables. Empty cells are now filled with NA. #232

tiny-plot

  • Class color assignment is now consistent across comparisons, and legend items are sorted alphabetically in the sample_avg_scatter_by_dge_class plot type. Color mapping now supports up to 20 distinct class colors (unless a stylesheet with >20 colors is provided by the user). #220
  • Plot outputs are now organized into subdirectories by type. #239

Full Changelog: v1.0.1...v1.1.0

Assets 2
Loading