Hi all,We have finally released a new version including several enhancements and bug fixes. As always, comments and suggestions are welcome. Also, we are developing a new documentation site, but progress is slow and we are now reworking the internals of bedtools to support future improvements and a better API for developer contributions.Thanks for all of the contributions so far and for using the tools!- AaronVersion 2.17.0 (3-Nov-2012)======================== New Tool ========================We have added a new tool (bedtools "jaccard") for measuring the Jaccard statisticbetween two interval files. The Jaccard stat measures the ratio of the lengthof the intersection over the length of the union of the two sets. In thiscase, the union is measured as the sum of the lengths of the intervals in eachset minus the length of the intersecting intervals. As such, the Jaccardstatistic provides a "distance" measure between 0 (no intersections)and 1 (self intersection). The higher the score, the more the two sets ofintervals overlap one another. This tool was motivated by Favorov et al, 2012.For more details, see see PMID: 22693437.We anticipate releasing other statistical measures in forthcoming releases.====================================== New Features & enhancements ======================================1. The genome file drives the BAM header in "bedtools bedtobam"2. Substantially improvement the performance of the -sorted option in"bedtools intersect" and "bedtools map". For many applications,bedtools is now nearly as fast as the BEDOPS suite when intersectingpre-sorted data. This improvement is thanks to Neil Kindlon, a staffscientist in the Quinlan lab.3. Tightened the logic for handling split (blocked) BAM and BED records4. Added ranged column selection to "bedtools groupby". Thanks to Brent Pedersen"- e.g., formerly "bedtools groupby -g 1,2,3,4,5"; now "-g 1-5"5. "bedtools getfasta" now properly extracts sequences based on blocked (BED12)records (e.g., exons from genes in BED12 format).6. "bedtools groupby" now allows a header line in the input.7. With -N, the user can now force the closest interval to have a different namefield in "bedtools closest"8. With -A, the user can now force the subtraction of entire interval whenany overlap exists in "bedtools subtract".9. "bedtools shuffle" can now shuffle BEDPE records.10. Improved random number generation.11. Added -split, -s, -S, -f, -r options to "bedtools multicov"12. Improvements to the regression testing framework.13. Standardized the tag reporting logic in "bedtools bamtobed"14. Improved the auto-detection of VCF format. Thanks to Michael James Clark.======================= Bug fixes =======================1. Fixed a bug in bedtobam's -bed12 mode.2. Properly include unaligned BAM alignments with "bedtools intersect"'s -v option.3. Fixed off by one error in "bedtools closest"'s -d option4."bedtools bamtobed" fails properly for non-existent file.5. Corrected missing tab in "bedtools annotate"'s header.6. Allow int or uint tags in "bedtools bamtobed"7. "bedtools flank" no longer attempts to take flanks prior to the start of achromosome.8. Eliminated an extraneous tab from "bedtools window" -c.9. Fixed a corner case in the -sorted algorithm.10.Prevent numeric overflow in "bedtools coverage -hist"
Sunday, 4 November 2012
Fwd: [bedtools-discuss] Version 2.17.0
Posted by Kevin at 10:20