v0.1.1 RC1

v0.1.1

• set number=1 when op=min see #40
• fix when we get -0 converted to just '-' (#42)
• when an annotation bed file has a header with 'ref' or 'reference' and 'alt' or 'alternate',
  then match on the ref and alt values defined in those columns (not just on position as for bed).
• (vcfgo) allow double-digit number in header thanks @chapmanb

v0.1.0

v0.1.0

This release contains usability improvements and some additional functionality.
Tar/zips below contain readme and examples. Just the binaries for linux-64 and osx 64 are also provided for ease of wgetting.

• drastically reduce the warning message output.
• add field=['DP2'] for bam files which reports number of forward and reverse reads overlapping a variant.
  value returned can be replicated with: samtools mpileup -A -x -Q 1 -q 1 -r chrom:pos-pos $bam
• add op=sum
• expose ref and alt to lua. (thanks @roryk)

v0.0.11

Version 0.0.11 contains a number of usability improvements along with fixes for the -ends stuff for annotating SVs. It should also have about 20% speed improvement thanks to the improvements in the go runtime itself.

v0.0.11

• when op=self, pull Number from the annotation file (previously Number was always 1)
• when op=concat or op=uniq user Number=.
• when name ends with '_float' '_int' or '_flag' that is used to determine the Type in the output and that is then stripped from the name. So, e.g. with
  names=["af_esp_aa_float"]
  The resulting header will be:

##INFO=<ID=af_esp_aa,Number=1,Type=Float,Description=...

• fix regression with CIPOS/CIEND when using -ends with structural variants. (Thanks to Liron for reporting)
• fix bug with uniq not uniq-ing (Thanks Liron for reporting)
• improve error messages from lua (Thanks Komal Rathi for the test-case provoking this change).
• improve handling of multiple values (usually due to multiple alts) for max and min. Now, if we have, e.g. var1:SAS=0.22,0.33
  and var2:SAS=0.25 then max will give 0.33. This change makes max and min recursive so they'll call themselves on nested slices.
  Thanks @snashraf for reporting.

0.0.10

In addition to performance and usability improvements, this release fixes some regressions. A functional test-suite has been expanded to improve these.

v0.0.10

• allow using postannotation even if not all requested fields were found for a given variant.
• restore ability to have bams as annotation files. Can pull mapq and coverage. See examples/conf.toml
• fix regression where output was not in sorted order.
• fix regression that resulted in "too many open files" error.
• expand test-suite.
• fix bug found when using max() op

v0.0.9

See below for full list of improvements. The biggest bug-fix is to fix a memory leak seen when annotating large files. The biggest change is to convert the scripting engine from javascript to lua. The reason for this was to improve the speed of user-defined functions. Now, functions defined in lua should be nearly as fast as the built-in functions but provide endless customization

v0.0.9

• restore ability to take query file from STDIN (no tabix required).
• fix memory leak. memory use now scales with number of procs (-p).
• added new op 'self' which should be used for most cases when matching on ref and alt as it
  determines the type from the annotation header and uses that to update the annotated header
  with the correct type.
• new documentation site
• [[postannotation]] allows modifying stuff in the query VCF after annotation (or instead).
  See examples on the documentation site.
• convert scripting engine to lua from javascript
• add CADD conversion script and example

parallel chromsweep.

This is a major change to vcfanno that allows use of parallel chromsweep. This removes several problems (namely chromosome ordering) with chrom-sweep. See this design document for more details.

The list of changes are:

v0.0.8

• parallel chrom-sweep (removes problems with chromosome sort order).
  • as a result, files are required to be tabix'ed.
  • the chromosome sort order is no longer important.
  • new -p flag indicates number of processes to use.
• fix bug in SV support of CIPOS, CIENDS
• huge speed improvement (can annotate ~30K variants/second with 10 cpus).
• remove server and cadd support (will return soon).
• fix bug where header is not updated.
• respect strict when -ends is used.

Thanks to @sigven for reporting several issues that arose as a result of the major code-changes.

server, speed, SVs

v0.0.7

• better support for flags. e.g. can specify a flag from js by ending the function name with _flag
• [irelate] error if intervals are out of order within a file.
• -base-path argument replaces basepath in .toml file
• [vcfgo] report all headers in original file.
• integrated server to host annotations
• -ends argument will now use CIPOS and CIEND to annotate the left and right interval of an SV. If CIPOS
  and CIEND are undefined for a given interval, the ends will not be annotated.
• for MNPs, cadd score is reported as a list of max values (of the 3 possible changes) for each reference base
  covered by the event.
• fix bug in CADD annotation and provide CADD v1.3 download
• ~25-30% speed improvement. from a modest laptop: *annotated 10195872 variants in 28.97 minutes (351984.0 / min

If you just want a linux binary without support files, it is here

CADD!

CADD is supported. Download encoded (11GB file) from:

https://s3.amazonaws.com/vcfanno/cadd_v1.2.bin
https://s3.amazonaws.com/vcfanno/cadd_v1.2.idx

and update path in conf

v0.0.6

• support for CADD
• concat defaults to | separator
• speed improvements (vcfgo info field)
• natural sort is default. use -lexographical to

v0.0.5

v0.0.5

• allow natural sort (1, 2, ... 9, 10, 11 instead of 1, 10, 11 ..., 19, 2, 20) via flag
• vcfgo: handle lines longer than 65KB major
• vcfgo: fix error reporting
• irelate: report warning when chroms out of order

v0.0.4

v0.0.4

• performance improvements for Javascript ops with pre-compilation.
• bam: annotate with mapq and seq for mapping-quality and sequence respectively.
• api now returns a channel on which to recieve annotated Relatables
• vcfgo: fix printing of INFO fields with multiple values (thanks to Liron for reporting).
• vcfgo: fix writing of ##SAMPLE and ##PEDIGREE headers. (thanks to Liron)