Releases: philres/ngmlr
Releases · philres/ngmlr
v0.2.7
- CMake fixes
- Dynamic linking is now default (use cmake -DSTATIC=ON for static
linking) --bam-fix
writes CS tag (Thanks Heng)- Fix to handle missing write permission when writing index
- Improved handling of smaller duplications
- Typos corrected (Thanks wdecoster)
- Support for RG in output
v0.2.6-beta
- Fixed: reads wrongly sometimes wrongly spanned two chromosomes (issues #23)
- Match score reduced and low quality split + short inv detection enabeld for -x ont
v0.2.5-beta
- Fix for long MD strings (segfault)
- Fix for LMP extension at chromosome borders (segfault)
- Dynamic linking is now default
v0.2.4-beta
- now reads from stdin if -q is skipped
- fixes for increasing stability
- updates and fixes for mapping ultra-long nanopore reads (https://github.com/nickloman/massive-nanopore-silliness)
- Parameter --bam-fix added. Currently, BAM does not support reads with > 64k cigar elements. When --bam-fix is set all reads with >64k Cigar elements will be reported as unmapped to maintain compatibility with BAM
v0.2.3-beta
v0.2.2-beta
- Compatibilty with (mini)conda
- SeqAn removed
- Switched from c++14 to c++11
- Option to switch off static build in CMake (-DSTATIC=OFF)
Mapping results should be the same as in v0.2.1
v0.2.1-beta
- Fixes for circular reference genomes
- Memory corruption fixed
- Filter for random short alignments added
- Dynamic programming approach for choosing best combination of mapped
segments (thanks to @bqminh) - Alignment computation speedup with SSE instructions (thanks to
@smolkmo) - Several memory leaks fixed
- Small duplication detection improved
--min-residues/-R
changed set to 25% (default). All reads that have
less than 25% mapped are reported as unmapped per default--max-segments
added. Max number of segments reported for a single
read
v0.1.6-beta
- Docker image for building ngmlr added
- Gap decay parameter added
- Memory corruption fixed
v0.1.5-beta
- MD tag fixed
Early beta version
0.1.4 Early beta release