GLIMPSE_phase (v1.1.1) segfault during finalization

[TedBrookings]

Hi, I'm getting a segfault similar to issue #92, where GLIMPSE_phase (v1.1.1) crashes during or after finalization. This only occurs with haploid data (in fact if I don't provide a --samples-file and thus run the sample for diploid imputation, it succeeds). I've pasted the log below:

GLIMPSE_phase --input $HOME/project/output/split-contig/Ind1-g_0.1x/chrX_Center.bcf --reference $HOME/project/output/split-phased-1kg/chrX_Center.bcf --impute-reference-only-variants --input-region chrX:2781514-5013923 --output-region chrX:2781514-4813922 --samples-file $HOME/project/output/split-contig/Ind1-g_0.1x/chrX_Center.sample-ploidy.tsv --output $HOME/project/output/glimpse1_impute/Ind1-g_0.1x/chrX_Center/00.bcf

[GLIMPSE] Phase and impute low coverage sequencing data
  * Author        : Simone RUBINACCI & Olivier DELANEAU, University of Lausanne
  * Contact       : [email protected] & [email protected]
  * Version       : 1.1.1
  * Run date      : 05/08/2024 - 10:14:31

Files:
  * Input VCF     : [$HOME/project/output/split-contig/downsampled/Ind1-g_0.1x/chrX_Center.bcf]
  * Reference VCF : [$HOME/project/output/split-phased-1kg/chrX_Center.bcf]
  * Output VCF    : [$HOME/project/output/glimpse1_impute/downsampled/Ind1-g_0.1x/chrX_Center/00.bcf]
  * Input region  : [chrX:2781514-5013923]
  * Output region : [chrX:2781514-4813922]

Parameters:
  * Seed       : 15052011
  * Input      : FORMAT/PL field for genotype likelihoods
  * #Threads   : 1
  * #Burnin    : 10
  * #Main      : 10
  * PBWT depth : 2
  * PBWT modulo: 8
  * HMM     : Constant recombination rate of 1cM per Mb
  * Ploidy     : given by samples file
  * Init K     : 1000

Initialization:

Parsing specified genomic regions
  * Input region  [chrX:2781514-5013923]
  * Output region  [chrX:2781514-4813922]
  * #samples = 1 [1 haploid/ 0 diploids]
  * VCF/BCF scanning [Reg=chrX:2781514-5013923 / L=39410 (Li= 1639 (4.15884%) - Lm= 37771 (95.8412%))] (0.23s)

WARNING: There are variants of the reference panel with no genotype likelihood in the target dataset. Please compute the likelihoods at all variants in the reference panel in order to use all the information of the sequencing reads.
  * VCF/BCF scanning [Nm=1 (Nmh=1 - Nmd=0) / Nr=3202 (Nrh=1598 - Nrd=1604)]
  * VCF/BCF parsing done (1.81s)
  * cM constant [1cM=1Mb] (0.00s)
  * Region spans 2232410 bp and 2.23 cM

Initializing iteration
  * HMM imputation [#states=1000.0 / %poly=70.4%] (0.30s)

Burn-in iteration [1/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.25s)
  * HMM imputation [#states=1000.0 / %poly=73.4%] (0.32s)

Burn-in iteration [2/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.28s)
  * HMM imputation [#states=1000.0 / %poly=73.0%] (0.28s)

Burn-in iteration [3/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.26s)
  * HMM imputation [#states=1000.0 / %poly=74.0%] (0.29s)

Burn-in iteration [4/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.26s)
  * HMM imputation [#states=1000.0 / %poly=74.3%] (0.28s)

Burn-in iteration [5/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.25s)
  * HMM imputation [#states=1000.0 / %poly=74.0%] (0.27s)

Burn-in iteration [6/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=73.6%] (0.29s)

Burn-in iteration [7/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=73.6%] (0.28s)

Burn-in iteration [8/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=73.2%] (0.27s)

Burn-in iteration [9/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.25s)
  * HMM imputation [#states=1000.0 / %poly=73.5%] (0.29s)

Burn-in iteration [10/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.25s)
  * HMM imputation [#states=1000.0 / %poly=73.8%] (0.32s)

Main iteration [1/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.27s)
  * HMM imputation [#states=1000.0 / %poly=72.6%] (0.29s)

Main iteration [2/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.25s)
  * HMM imputation [#states=1000.0 / %poly=73.5%] (0.38s)

Main iteration [3/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.25s)
  * HMM imputation [#states=1000.0 / %poly=73.9%] (0.29s)

Main iteration [4/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.25s)
  * HMM imputation [#states=1000.0 / %poly=73.4%] (0.27s)

Main iteration [5/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=73.3%] (0.27s)

Main iteration [6/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=72.5%] (0.27s)

Main iteration [7/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=72.4%] (0.27s)

Main iteration [8/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=73.6%] (0.27s)

Main iteration [9/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=73.0%] (0.27s)

Main iteration [10/10]
  * HAP update (0.00s)
  * PBWT building & selection (0.24s)
  * HMM imputation [#states=1000.0 / %poly=72.8%] (0.27s)

Finalization
  * done

Finalization:
  * BCF writing [Compressed / N=1 / L=39410] (0.08s)
  * Total running time = 13 seconds
/bin/bash: line 20: 90687 Segmentation fault: 11  GLIMPSE_phase --input "$HOME/project/output/split-contig/downsampled/Ind1-g_0.1x/chrX_Center.bcf" --reference "$HOME/project/output/split-phased-1kg/chrX_Center.bcf" --impute-reference-only-variants --input-region "$input_region" --output-region "$output_region" --samples-file "$HOME/project/output/split-contig/downsampled/Ind1-g_0.1x/chrX_Center.sample-ploidy.tsv" --output "$out_bcf"

[srubinacci]

Hi,
It seems the the program runs fine, but it has some deallocation error at the end of the code. Do you have a .BCF output out of this run?