final vcf file

[AnnaAlk]

Can you please explain the values in the final vcf file, especially for ALT and INFO columns?

#CHROM POS ID REF ALT QUAL FILTER INFO
1 726104 1:726104:FP A A[COMPONENT_947_a_length_353_size_5f:135[ . . AR=2;AS=0.5;GS=1
1 725161 1:725161:FP G ]COMPONENT_947_a_length_353_size_5f:277]G . . NOANCHOR

[bkehr]

The ALT field follows the breakend notation specified in the VCF4.2 standard, section 5.4 and specifically section 5.4.2. The contig names refer to the sequences in supercontigs.fa. If the name is followed by an f, the sequence is inserted in the current orientation. If it is followed by an r, it is inserted in reverse complemented orientation and the position refers to the reverse complemented sequence.

Contig names in supercontigs.fa start with COMPONENT if they have been merged from contigs of several individuals. Component names indicate the length of the contig and the number of merged contigs (size). If the component is branching (Kehr et al., Bioinformatics 2016), the lowercase letter between the component number and length distinguishes the different paths through the graph. If a contig has not been merged, the the contig names output by Velvet are used and prefixed with a sample ID.

The fields in the INFO column should be defined in the header of the VCF file, but they currently aren't:
AR is the number of anchoring reads, AS is the anchoring score, and GS is the group size (Section "Popins placing step" in Supplementary text of Kehr et al., Nat Genet 2017). The tag NOANCHOR indicates that no anchoring read pair was present and the location was tested for split alignments based on the anchoring of the other end of the contig. Records tagged with NOANCHOR are generally less reliable.

Thanks for reporting this, we will add the definition of the INFO fields to the output VCF header and a short documentation of the output file to the README.