"Variant is not consistent: U00096.3:965-968 - TGCT(REF) vs (FASTA)" error

[VadimDu]

Hello @tseemann and all users,

I am running snippy with basic configurations and keep getting the following error (even when change the reference from FASTA to GenBank):

[variant_manip.cpp:75 is_ref_consistent] Variant is not consistent: U00096.3:965-968 - TGCT(REF) vs (FASTA) [normalize.cpp:178 normalize] Normalization not performed due to inconsistent reference sequences. (use -n option to relax this) [main_vcfview] Error: cannot write to (null)

The result is always "Identified 0 variants", needless to say that the target species is present in high abundance in the sample. I might be missing something, any help is much appreciated.

Thank you very for the assistance, Best regards Vadim

[fengyuchengdu]

[variant_manip.cpp:75 is_ref_consistent] Variant is not consistent: SCKP140071_chr:355431-355441 - GAAAAAAAGTC(REF) vs GATCCTGCTGG(FASTA) [normalize.cpp:178 normalize] Normalization not performed due to inconsistent reference sequences. (use -n option to relax this)

Pretty much the same. Never had this error before.

[VadimDu]

Hi @fengyuchengdu

Thanks for your replay. So I understand that you did use snippy before, on the same data? and never got this error? maybe there was some version changes in one of the dependencies.

[fengyuchengdu]

Hi @VadimDu After some tests, the error seems to arise from the recent update (2015.11.10-h5ef6573_4) of dependency vt. I solved the problem by rolling vt back to version 2015.11.10-he941832_3.

[eam12]

Just updated to 4.6.0 and now I'm getting the same error for a bunch of samples that ran perfectly before...

[variant_manip.cpp:75 is_ref_consistent] Variant is not consistent: CP016412.1:973613-973615 - GCG(REF) vs CGC(FASTA) [normalize.cpp:178 normalize] Normalization not performed due to inconsistent reference sequences. (use -n option to relax this) [main_vcfview] Error: cannot write to (null)

[VadimDu]

Hi @eam12

I was managed to fix this error thanks to the solution by @ktmeaton from issue #410 (by downgrading dependency [vt]) - look there.

[eam12]

Thanks, @VadimDu! Seems to be working now.

[wanyuac]

I had the same issue with @VadimDu and solved the problem as @fengyuchengdu has suggested. Specifically, my Snippy version is v4.6.0, installed through anaconda. The default version of vt was 2015.11.10 h5ef6573_4. The vt command compianed about inconsistent reference. As a result, the following command in the snippy pipeline produced an empty VCF and did not call any variant eventually (a severe problem of false negatives!)

bcftools view --include 'FMT/GT="1/1" && QUAL>=100 && FMT/DP>=10 && (FMT/AO)/(FMT/DP)>=0' snps.raw.vcf  | vt normalize -r reference/ref.fa - | bcftools annotate --remove '^INFO/TYPE,^INFO/DP,^INFO/RO,^INFO/AO,^INFO/AB,^FORMAT/GT,^FORMAT/DP,^FORMAT/RO,^FORMAT/AO,^FORMAT/QR,^FORMAT/QA,^FORMAT/GL' > snps_filt_manual.vcf

Command lines I have used to solve this problem were copied from Issue 410.

conda create --name snippy2
conda install --name snippy2 -c bioconda/label/cf201901 vt==2015.11.10  # vt 2015.11.10 he941832_3 bioconda/label/cf201901
conda install --name snippy2 -c bioconda snippy

Of note, we cannot use the option -n of vt normalize to circumvent this issue as indels will be skipped and not be normalised in this way, which goes against the purpose of using vt in the snippy pipeline.

[andersgs]

This is a known bug that we are in process of fixing. In the mean time, we do recommend downgrading. Thank you.

[AmanuelZewdu]

hi dear How can i downgrade the dependency vi?

[AmanuelZewdu]

Thanks, @VadimDu! Seems to be working now.

hi dear how can i downgrade the dependency vi?