v0.9.5

This release adds new features and fix some minor bugs

The addition of the new features required to modify and update the germline pre-processing scripts to implement dragen VCFs pre-processings

New Features

• processing VCF from two new tools: dragen_snv and dragen_sv ; These tools are part of the Illumina Dragen Pipeline ;
• remove duplicated lines from dragen_sv tool using an external tool implemented in python3 (main_dedup_sv.py ; needs to be in the PATH)

Bugs fixed

• fix keyError bug: replace data[0] with data[tool]
• replace cp with rsync
• check the step value that update the counter is correctly assigned when the total count of variant from the union of all the VCF is less than 10 variants
• Fix indels filtering out with Phaser; Encountered edge cases; fix issue #27.

v0.9.4

Fix

• Issue #27 : Added steps to remove first the Indels from the VCF because phASER does not phase indels and that can cause issue if a SNV and an Indel are found in consecutive positions. When so phASER fails because after excluding indel, phASER remains with only one line SNV and cannot phase a unique line; phasing requests at least 2 lines of consecutive SNV
• Issue #26 : updated text that better explain what is performed with hom and hets

v0.9.3

Bug Fix version

• Octopus vcf prep bug fix
• Wrong Value assignment to ARs field [Allelic Ratio]; Fix to correct value when grabbing the value from the original AF field generated by Octopus tool. Because AR was always set to zero, all the variants got filtered out in the Phoenix/Tempe workflows.
• Note: AR value for normal sample is 0 or 1; When one, it is often due to multiallelic sites;

v0.9.2

Improvement after Error with phASER tool

• added check for testing if there are HETS calls in the vcf input for phASER tool;
• Note_1: If no hets (which is different from if no calls), we skip running phASER; Previous check was not thorough enough since it only checked if the vcf contained at least 2 calls; but in the edage case we got only indel in the VCF and phASEr failed because of that case and by returning an exit value of 1, the entire pipeline for vcfMerger2 failed.
• Note_2: This case had never been encounter before from the thousands of samples we ran.

v0.9.1

Bug Fix

• Fix issue with pop(0) by adding checkpoint if dictionary is empty

Improvements

• added debug lines
• improved --debug usage by using the getLogger() method

v0.9.0

This release is only due to the tool octopus for which the vcf format got modified a lot from the previous octopus versions (i.e. less than v.0.7.4) and got not compatible with the vcfMerger2 version 0.8.4 anymore.

major changes:

• this version is not backward compatible with vcfMerger2 releases less or equal to 0.8.4.
• this version is compatible with octopus version 0.7.4 or up (if octopus is used obviously :-) )
• modifications in prep steps for octopus vcf file, where the vcf file has been generated by the octopus version 0.7.4 or up which follows VCF v4.3 specs.

minor changes:

• cleaner code for the scripts and files modified for making that version of vcfMerger2: mostly removing spaces, typos, indentations, log.info or log.debug information.
• capture AR field info from the AF field created by octopus itself and not by recalculating AR from recaptured AD information.
• added a step in the octopus prep vcf after vt decompose to remove the stars that are adding new information (see VCF specs v4.3) to the call. vt tool could not interpret the * and did not de-convolute properly the lines with a star. That step got added to the octopus prep remove_star_from_octopus_decomposed_vcf.
• GenotypeInv class got updated to handle upto 4 alleles in the genotype ; if more than four (4) alleles exist in the genotype of one of the two samples, octopus prep steps will ignore the 5th or more allele; Either user must filter out the calls with 5 alleles or more or make a pull request or raise an issue in the vcfrMerger2 github repo.
• MAP_VAF and VAF_CR got renamed MAP_HF and HF_CR in the vcf of version 0.7.4 of octopus. see commit: (luntergroup/octopus@48e2b5b) .

v0.8.7

update/FIX

• Updated the Genotype Class to capture up to 9 genotypes values in the GT field; was 6 before except for Octopus which was 3 only.

v0.8.6

Fix

• remove check version for bcftools; did not work when bcftools version has two digits in the decimal version part

v0.8.5

fix/update

• fix issue #18
• updated the flag TYPE in the merged vcf to VTYPE in order to avoid conflict with bcftools which uses TYPE as a special keyword with specific values associated.
  See URL for further explanation about TYPE in bcftools : http://samtools.github.io/bcftools/bcftools.html#expressions

v0.8.4

Fix Edge Case Octopus