fix regression from 0.3.4

This fixes a major bug in v0.3.4 (#150). Please update.

Full Changelog: v0.3.4...v0.3.5

v0.3.4

This release fixes and issue where the FORMAT column was added to the VCF header when no samples were present.

Full Changelog: v0.3.3...v0.3.4

v0.3.3

This contains minor usability fixes.

• 0314e53 - add vcfanno version to header (#140)
• c4c0ae9 - update travis
• 49ef500 - non-zero exit when no args are given
• 6e5a12a - fix help which still mentioned js.
• 9f857a9 - handle missing value in postannotation. closes #116
• 1d140d4 - Misleading Flatten error message (#119)
• 8ffa89c - fix example table format of markdown (#115)

general fixes and by_alt

this release has minor fixes to by_alt and other bugs:

• 8c692e2 - handle by_alt + float correctly (#113) (12 days ago)
• deeeee3 - Fix always falling back to non-bgzip reader (#109) (8 weeks ago)
• 54bcc00 - add go 1.11 (5 months ago)
• cc8048d - add release target to makefile (5 months ago)
• 5683246 - Add Makefile (#103) (5 months ago)
• 16701a1 - closes #99 (6 months ago)

fix for empty description

v0.3.1

• allow empty description in VCF header (sigven/pcgr#49)

minor updates

v0.3.0

• support for NON_REF to match any ALT allele (#92)
• more work on multiple alts (#87)

there should be ~3-5% speed improvement thanks to improvements in the go compiler.

csi support

this release adds support for CSI. this can greatly speed up dense files like CADD and even gnomAD or ExAC.

to use, simply index your file with tabix as, e.g.:

tabix -m 12 --csi $file

and vcfanno will prefer the .csi index over the .tbi.

refer to the performance page for how to optimize.
in brief, setting

export IRELATE_MAX_GAP=1000

seems to give good results. Please experiment and let me know what works for you. I can get ~60% speed improvement using csi'd ExAC to annotate clinvar with the above changes.

bug fix

v0.2.8

• fix fo #71 for out of bounds warning followed by panic.

improvements for multiple alts (self, by_alt ops)

v0.2.7

• restore multiple threads per annotation file after fix in biogo/hts (#64)
• new op setid builtin to set the ID field and remove need for lua and fix some bugs.
• fix bug in self with multiple alternates when there were multiple overlaps (thanks Matthew).
• fix for #68, #69 -- with Number=A, op="self", the output will always have a number of elements
  equal to the number of alternates. If there are > 1 annos for a given site, later values will
  overwrite previous.
• new op by_alt that will have Number=A and will append multiple annotations for the same alternate
  (from the same file) and output them as pipe-delimited. e.g with 2 alts, it might look like: 0.111,0.222|0.333
  when the 1st alternate has a single value and the 2nd alternate has 2 values.

by_alt and other multi-allelic fixes

v0.2.7 (dev)

• restore multiple threads per annotation file after fix in biogo/hts (#64)
• new op setid builtin to set the ID field and remove need for lua and fix some bugs.
• fix bug in self with multiple alternates when there were multiple overlaps (thanks Matthew).
• fix for #68, #69 -- with Number=A, op="self", the output will always have a number of elements
  equal to the number of alternates. If there are > 1 annos for a given site, later values will
  overwrite previous.
• new op by_alt that will have Number=A and will append multiple annotations for the same alternate
  (from the same file) and output them as pipe-delimited. e.g with 2 alts, it might look like: 0.111,0.222|0.333
  when the 1st alternate has a single value and the 2nd alternate has 2 values.