Releases · richardmleggett/nextclip

Changes since 0.9:

Bug fixes

Improved ability to spot junction adaptor due to bug in previous logic.
Removed some crashes that can occur if some of the stats counts are zero (ie. dividing zero by number of reads creates error).
Corrected LSF submission to request correct amount of memory for NextClip (and provided option which will report memory requirements before running NextClip).
The min_length parameter was 1 base out (i.e. specifying 25 gave reads of 24) - this is fixed, meaning that yields will appear to go down slightly. You can always decrease min_length to 24 if you want to recover this.
Fixed PCR duplicate stats being wrong for counts > 99.
Fixed help wording.

Improvements

NextClip outputs number of bases written for each category and the pipeline PDF report shows this as well as approximate genome coverage.
Ability to pass parameters through to the pipeline (e.g. minimum width, trim ends etc.).
You can specify the pipeline stage to start at, so no need to redo all analysis in the event of a fail.
The scripts directory is found automatically - no need to set it during installation.
Moved PCR duplicate kmer offset to avoid lower quality bases at start of read. Should provide slightly more accurate estimates of PCR duplication.
Queue name can be specified when running with LSF.
Reduced size of hash table used to store PCR duplicates.

Provide feedback