Quilt2 Impute

[PMuchina]

I am trying out Quilt2 and I ran into this.

[2024-09-30 11:38:12] i_gibbs=1, i_it = 1 full
Error in sample.int(x, size, replace, prob) :
cannot take a sample larger than the population when 'replace = FALSE'
Calls: QUILT ... select_new_haps_mspbwt_v3 -> unique -> sample -> sample.int
Execution halted

[Zilong-Li]

Thanks for reporting this.

Can you provide more information? say the command line, the reference panel and samples as well as the R environment by sessionInfo()

[PMuchina]

I want to impute 30 samples

QUILT2.R
--prepared_reference_filename=QUILT_prepared_reference.Chr1.104936818.109936793.RData
--bamlist=/1x/bamlist
--method=diploid
--chr=Chr1
--regionStart=104936818
--regionEnd=109936793
--buffer=500000
--output_filename=quilt2_output/quilt2.Chr1.104936818.109936793.vcf.gz

sessionInfo()
R version 4.3.3 (2024-02-29)
Platform: x86_64-conda-linux-gnu (64-bit)
Running under: Debian GNU/Linux 12 (bookworm)

Matrix products: default
BLAS/LAPACK: /usr/home/qgg/pekmu/miniconda3/envs/Impute/lib/libopenblasp-r0.3.27.so; LAPACK version 3.12.0

locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
[7] LC_PAPER=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

time zone: Europe/Copenhagen
tzcode source: system (glibc)

attached base packages:
[1] stats graphics grDevices utils datasets methods base

loaded via a namespace (and not attached):
[1] compiler_4.3.3

[Zilong-Li]

may I ask how many haplotypes in the referencce panel? and could you do library(QUILT) first and then show sessionInfo()

[PMuchina]

274 haplotypes

[2024-09-30 13:51:38] Auto-set n_burn_in_seek_its to 2 i.e. only sample one dosage per Gibbs sample
[2024-09-30 13:51:38] Overriding default parameters for small reference panel
[2024-09-30 13:51:38] Observing number of reference haplotypes K=274
[2024-09-30 13:51:38] Reset n_seek_its from 3 to 1
[2024-09-30 13:51:38] Reset n_burn_in_seek_its from 2 to 0
[2024-09-30 13:51:38] Set Ksubset to 274 (no longer necessary)
[2024-09-30 13:51:38] Set Knew to 274 (no longer necessary)
[2024-09-30 13:51:38] Get BAM sample names
[2024-09-30 13:51:39] Done getting BAM sample names
[2024-09-30 13:51:39] There are 171352 common SNPs and 186842 SNPs overall in this region
[2024-09-30 13:51:40] There are 0 regions out of 5838 below minimum recombination rate, setting them to minimum rate
[2024-09-30 13:51:40] There are 0 regions out of 5838 above maximum recombination rate, setting them to maximum rate
[2024-09-30 13:51:40] Imputing sample: 1
[2024-09-30 13:51:41] The average depth of this sample is:0.970354809598658
[2024-09-30 13:51:41] There are 12257 reads under consideration
[2024-09-30 13:51:41] i_gibbs=1, i_it = 1 small gibbs
[2024-09-30 13:51:42] i_gibbs=1, i_it = 1 full

[PMuchina]

library(QUILT)
sessionInfo()
R version 4.3.3 (2024-02-29)
Platform: x86_64-conda-linux-gnu (64-bit)
Running under: Debian GNU/Linux 12 (bookworm)

Matrix products: default
BLAS/LAPACK: /usr/home/qgg/pekmu/miniconda3/envs/Impute/lib/libopenblasp-r0.3.27.so; LAPACK version 3.12.0

locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
[7] LC_PAPER=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

time zone: Europe/Copenhagen
tzcode source: system (glibc)

attached base packages:
[1] parallel stats graphics grDevices utils datasets methods
[8] base

other attached packages:
[1] QUILT_2.0.0 data.table_1.16.0 mspbwt_0.1.0 STITCH_1.7.1

loaded via a namespace (and not attached):
[1] compiler_4.3.3 Rcpp_1.0.13

[Zilong-Li]

Thanks. I see two issues:

1. the number of haplotypes (274) is smaller than the default (600). Can you try explicitly setting --Ksubset=274 and --Knew=274. We'll fix this in the near future.
2. QUILT has some issues with Rcpp==1.0.13. It's recommended to use Rcpp==1.0.12 or the bioconda.

[PMuchina]

I am still facing the same error.

But I just faced this with Quilt2. Quilt 1 runs well.