search

bioinformatics-centre / BayesTyper

A method for variant graph genotyping based on exact alignment of k-mers
87 stars 7 forks source link

bayestyper cluster assertion failure #17

Closed cheesemania closed 5 years ago

cheesemania commented 5 years ago

I'm running into an error with the cluster tool

[08/08/2019 16:08:15] You are using BayesTyper (v1.4)

[08/08/2019 16:08:15] Seeding pseudo-random number generator with 1565298495 ... [08/08/2019 16:08:15] Setting the kmer size to 55 ...

[08/08/2019 16:08:15] Parsed information for 1 sample(s)

[08/08/2019 16:08:15] Parsing reference genome ... [08/08/2019 16:08:16] Parsed 14 reference genome chromosomes(s) (24214674 nucleotides)

[08/08/2019 16:08:16] Parsing decoy sequence(s) ... [08/08/2019 16:08:16] Parsed 3 decoy sequence(s) (4837922 nucleotides)

[08/08/2019 16:08:16] Setting the number of inference units to 1 across 951939 variants ...

[08/08/2019 16:08:23] Maximum resident set size: 1.80473 Gb

[08/08/2019 16:08:23] Parsing variants in unit 1 ... bayesTyper: /isdata/kroghgrp/jasi/bayesTyper/code/releases/v1.4_static/BayesTyper-1.4/src/bayesTyper/VariantFileParser.cpp:218: bool VariantFileParser::constructVariantClusterGroups(InferenceUnit*, uint, const Chromosomes&): Assertion `num_variant_clusters_pre < num_variant_clusters' failed. ./variantcluster.sh: line 8: 14522 Aborted /master/ianc/BayesTyper/bayesTyper_v1.4_linux_x86_64/bin/bayesTyper cluster -p 8 -v candidates.vcf.gz -s ${1}.tsv -g PvivaxP01.genome_canon_2.fa -d PvivaxP01.genome_decoy_2.fa -o ${1}_bayestyper

I'm not certain whats going wrong, as near as I can tell the variants file looks good and has been sorted. Any clues about whats happening here?

Many Thanks,

Ian

cheesemania commented 5 years ago

UPDATE

I have upgraded to version 1.5 and no longer get this error message

jonassibbesen commented 5 years ago

Glad to hear that it worked with the new version. This error normally arises when BayesTyper is not able to properly parse the vcf file. Do the reference sequences in your vcf file have a different case (upper vs. lower) than the input reference genome? This is the only relevant issue that have been fixed in the vcf parsing code since v1.4. I am trying to understand why you got this error in the first place in order to make sure that the issue is completely fixed.

Please let me know if you have any other questions or issues.

Thanks,

Jonas