Ataqv Versions Save

• Fix missing TSS enrichment when using --ignore-read-groups

• Add option to use single nucleus barcode tag

• Use default read group during TSS enrichment calculation if read doesn't have RG tag

Fix parsing of empty BAM header tags.

Add option to reduce redundant metrics.

Add HQAA chromosome read counts (in JSON only). Add max fraction autosomal reads from single chromosome metric.

Speed up TSS coverage/enrichment calculations when multiple read groups are present. Also, TSS coverage/enrichment was previously (incorrectly) the same for all read groups in the bam file. That's no longer the case.

Fix omitted metrics bug introduced with concurrency. If an ataqv output file contained more than one set of metrics, only the first would be added to the viewer.

Add table export to CSV.

Add fly, worm, yeast autosomes.

With hundreds of libraries, the viewer could bog down. This release attempts to address that. The visible changes are that plot lines and dots are now opaque (translucence meant compositing slowdowns) and that the experiment detail tooltip only appears when plot items are clicked.

You can now filter the plot items by searching the text of their identifiers and descriptions.

Changes to get it to build and test on Macs, or on Linux with HTSlib 1.4 or newer.

Macs have /usr/bin/true, not /bin/true.

HTSlib has added JSON support, which introduced a namespace clash, and data URLs passed to sam_open now require a comma before the content; the old HTSlib format has been kicked to the curb.

Only consider autosomal alignments for peak duplicate metrics.

You can now export plots to SVG and PNG, but the exports are only accurate in Firefox and Chrome so far.

Correct maximum fragment length observation

• The code didn't match its comment; when trying to ascertain the maximum proper fragment length, it was considering supplementary (chimeric) and secondary alignments. Corrected it to only measure the fragment size of primary alignments.

New metrics: duplicates in and out of peaks

• Measure the fraction of duplicate reads outside of and within peaks.
• Refactored some measurements, moving them from Metrics to Peaks.
• Purged a bunch of YAGNI code.
• Brought test coverage up to date.

Instead of counting fragment lengths of any properly-paired and mapped reads, only count high-quality autosomal alignments. This makes the fragment length distribution plot more consistent with the others, and often shows nucleosomal periodicity better.

This release also includes improved extents and zooming for the TSS enrichment plot.