header format of .tsv did not match! about ivar

[Alex-Vasile]

We also ran into this issue and dug into a bit. It's caused by the extra POS_AA column at the end.

Temporary solution for anyone having this issue

If you don't need the POS_AA data column, pre-process your variant files to remove this column.

In-depth Info

1. call_variants_from_plup prints out a POS_AA (from a hardcoded set of column headings inside the function).
   

   ivar/src/call_variants.cpp

   Lines 58 to 61 in 08aac33

   	"\tALT_CODON"
   	"\tALT_AA"
   	"\tPOS_AA"
   	<< std::endl;

2. common_variants calls read_variant_file which first checks if the headers are correct:
   

   ivar/src/get_common_variants.cpp

   Lines 25 to 30 in 08aac33

   	while (std::getline(line_stream, cell, '\t')) {
   	if (cell.compare(fields[ctr]) != 0) {
   	return -1;
   	}
   	ctr++;
   	}

However this checks has 2 issues with it:

1. It uses a parallel, and out of date, set of header names; it's missing POS_AA. This and call_variants_from_plup should be working with a single set of fields so there aren't two parallel structures to update when a change happens.
2. This code is has an out of bounds error, which is what's happening now. The loop will keep reading heading columns and index into fields even after ctr >= NUM_FIELDS. The loop should terminate if there are more than NUM_FIELDS entries.

ivar/src/get_common_variants.cpp

Lines 3 to 8 in 08aac33

	const int NUM_FIELDS = 19;
	const std::string fields[NUM_FIELDS] = {
	"REGION", "POS", "REF", "ALT", "REF_DP",
	"REF_RV", "REF_QUAL", "ALT_DP", "ALT_RV", "ALT_QUAL",
	"ALT_FREQ", "TOTAL_DP", "PVAL", "PASS", "GFF_FEATURE",
	"REF_CODON", "REF_AA", "ALT_CODON", "ALT_AA"};

Also worth considering is changing the error message from common_variants. It currently gives the incorrect impression that the header formats of A_variant and B_variant do not match each other, but what it actually means is that they don't match the expected header. Would be worth changing that message and also printing both the received header and the expected header.

[cmaceves]

Hi! Sorry that you're having issues, thanks for reaching out. Could you be more specific about the origin of "Part 4" and maybe share the .tsv files with me? Based on the given error, I would assume that the SNV variants files being used are not properly formatted but it's hard to tell without sample files!

[zdk427]

Sure here is the link to .tsv files i got after process 3 in folder 6.Singlet-SNVs
Please let me know if you need any further information.
https://usaskca1-my.sharepoint.com/:f:/g/personal/zdk427_usask_ca/EolxZ7UcEDpPo6qFVXKB_KoBajXB_8WyJ6k5Kx3QoU3OjA?e=Y1ihRt