I've downloaded the chr X or 1000g reference again because there is no chrX in the folder downloaded from https://imputationserver.sph.umich.edu/static/downloads/releases/1000genomes-phase3-2.0.0.zip.
Nevertheless, I found some issues for imputation of chrX using docker image of Michigan Imutation Server.
-
The sever cannot find reference information of chrX.
[ERROR] Minimac reference panel cloudgene/apps/1000g-phase-3-v5@2.0.0/2.0.0/m3vcfs/X.nonPAR.1000g.Phase3.v5.With.Parameter.Estimates.m3vcf.gz not found.
Job chr_X.nonPAR (null) failed.
-
Wrong matching between my dataset and reference for chrX
I've already checked my data using 1000 genome reference before imputation (https://www.well.ox.ac.uk/~wrayner/tools/).
I uploaded QCed input files.
I used "EAS" population and selected the same population of 1000 genome data for the imputation.
However, the qcreport reported wrong information.
Most of mismatched frequencies were for chrX.
I think the server did not use the chrX of EAS population.
I checked the frequencies of the variants reported in the qcreport, but the frequences were not those of EAS population.
I attached my qcreport and log files from the docker server.
Please check them.
errors.zip
I've downloaded the chr X or 1000g reference again because there is no chrX in the folder downloaded from https://imputationserver.sph.umich.edu/static/downloads/releases/1000genomes-phase3-2.0.0.zip.
Nevertheless, I found some issues for imputation of chrX using docker image of Michigan Imutation Server.
The sever cannot find reference information of chrX.
[ERROR] Minimac reference panel cloudgene/apps/1000g-phase-3-v5@2.0.0/2.0.0/m3vcfs/X.nonPAR.1000g.Phase3.v5.With.Parameter.Estimates.m3vcf.gz not found.
Job chr_X.nonPAR (null) failed.
Wrong matching between my dataset and reference for chrX
I've already checked my data using 1000 genome reference before imputation (https://www.well.ox.ac.uk/~wrayner/tools/).
I uploaded QCed input files.
I used "EAS" population and selected the same population of 1000 genome data for the imputation.
However, the qcreport reported wrong information.
Most of mismatched frequencies were for chrX.
I think the server did not use the chrX of EAS population.
I checked the frequencies of the variants reported in the qcreport, but the frequences were not those of EAS population.
I attached my qcreport and log files from the docker server.
Please check them.
errors.zip