NEWS

ORFik 1.23.8

Fixed allow.introns forwarding in STAR.align.single and STAR.align.folder. The default TRUE now enables novel junction discovery (--alignIntronMax 0). Earlier affected runs silently used 1 even with TRUE. Set FALSE explicitly to retain that indexed-junction-only behavior; indexed splicing remains possible in either mode. This correction can change alignment results.
Added base.correction = FALSE to both entry points. TRUE enables fastp correction of overlapping paired reads during trimming (issue #172).
Pipeline logs now show each step's actual inputs, outputs and command, instead of repeatedly presenting the original reads as input (issue #174).
Fixed paired-folder multimapping option forwarding, tRNA input compression, and noShared memory mode across multiple libraries. Commands preserve paths containing spaces or shell metacharacters and stop on tool failure.
Empty folders, duplicate output basenames, and invalid step order now fail explicitly. Pairing still uses adjacent files in sorted filename order.

Added pmapToTranscriptsF, a much faster pmapToTranscripts.
All of ORFik now supports weights for representing duplicated reads. This will speed up all function related to NGS data a lot.

mapToGRanges is now much faster and uses much less memory. A c++ reimplementation of pmapFromTranscripts from GenomicFeatures.

Remade many function to handle no names correctly, like tile1 will now work even for GRangesList without names. Documentation is more logical, and removed coverageByWindow, since GenomicFeatures::coverageByTranscripts now works correctly.
The orf finding function now find the longest orf per stop codon if you set longestORF = TRUE in findORFS, findMapORFs and findORFsFasta

first release of ORFik - find Open Reading Frames, automatic RiboSeq footprint shifts, reassignment of Transcription Start Sites with the use of CageSeq, plethora of gene identity functions from scientific publications Bioconductor build fixes
Genome download failures now report the underlying service error instead of failing while coercing an error object to a logical value.
The importing-data and transcript tutorials use bundled synthetic references for runnable examples; network download examples remain available as explicit, optional steps. All vignettes include session info.
Reference-space CIGAR widths now use cigarillo on current Bioconductor, with a compatibility fallback for older releases, as query-space widths do.
Clarified local-reference arguments, cache reuse and pseudo-leader interpretation, and corrected the fastp UMI example.