search

statgen / METAL

Meta-analysis of genomewide association scans
Other
39 stars 12 forks source link

METAL HAS CRASHED, but the issue does not seem to be related to memory access. #33

Closed 1667857557 closed 2 weeks ago

1667857557 commented 1 year ago

Dear researcher, Thank you for your work. Metal is a great tool in GWAS meta-analysis. However, when We including a summary GWAS file with over 800 million lines (an 846M file), it reports an error. "_ETAL HAS CRASHED

The operating system has decided to terminate METAL. This could be due to a memory access problem, or it could be you encountered a bug.

To help improve this program, please report bugs by e-mailing goncalo@umich.edu. Please include details of how the problem you encountered can be reproduced_."

We checked the system memory, and it was sufficiently full to perform this tool. Additionally, we attempted to switch to another computer, but encountered the same error. here is our input file header : "SNP CHR BP A1 A2 BETA P SE FRQ N rs575272151 1 11008 G C -0.0303280157171869 0.0413221 0.014864608696787 0.0943995 247173 rs544419019 1 11012 G C -0.0303279257228457 0.0413229 0.0148646206629449 0.0943995 247173 rs62635286 1 13116 G T 0.0116022593832627 0.387424 0.0134238973205239 0.175915 247173 rs62028691 1 13118 G A 0.0116024132307965 0.387417 0.0134238786313126 0.175915 247173"

and

_"MARKER SNP WEIGHT N ALLELE A1 A2 EFFECT BETA STDERR SE FREQ FRQ PVAL P

PROCESS All_Breast_cancer.txt

SCHEME STDERR
STDERR SE
OVERLAP ON
OUTFILE BC .TBL GENOMICCONTROL ON
MINMAXFREQ ON
AVERAGEFREQ ON
ANALYZE Quit"_ looking forward to your reply###

1667857557 commented 1 year ago

Dear researcher, Thank you for your work. Metal is a great tool in GWAS meta-analysis. However, when We including a summary GWAS file with over 800 million lines (an 846M file), it reports an error. "_ETAL HAS CRASHED

The operating system has decided to terminate METAL. This could be due to a memory access problem, or it could be you encountered a bug.

To help improve this program, please report bugs by e-mailing goncalo@umich.edu. Please include details of how the problem you encountered can be reproduced_."

We checked the system memory, and it was sufficiently full to perform this tool. Additionally, we attempted to switch to another computer, but encountered the same error. here is our input file header : "SNP CHR BP A1 A2 BETA P SE FRQ N rs575272151 1 11008 G C -0.0303280157171869 0.0413221 0.014864608696787 0.0943995 247173 rs544419019 1 11012 G C -0.0303279257228457 0.0413229 0.0148646206629449 0.0943995 247173 rs62635286 1 13116 G T 0.0116022593832627 0.387424 0.0134238973205239 0.175915 247173 rs62028691 1 13118 G A 0.0116024132307965 0.387417 0.0134238786313126 0.175915 247173"

and

_"MARKER SNP WEIGHT N ALLELE A1 A2 EFFECT BETA STDERR SE FREQ FRQ PVAL P

PROCESS All_Breast_cancer.txt

SCHEME STDERR STDERR SE OVERLAP ON OUTFILE BC .TBL GENOMICCONTROL ON MINMAXFREQ ON AVERAGEFREQ ON ANALYZE Quit"_ looking forward to your reply###

This error only exist in window system, METAL perform well in linux

abecasis commented 1 year ago

I think to solve the problem you probably need to download the Metal source and recompile. The old binaries can probably only address 2-4 GB of RAM.

G

On Thu, Jun 15, 2023 at 8:50 PM Yu-Feng Huang @.***> wrote:

Dear researcher, Thank you for your work. Metal is a great tool in GWAS meta-analysis. However, when We including a summary GWAS file with over 800 million lines (an 846M file), it reports an error. "_ETAL HAS CRASHED

The operating system has decided to terminate METAL. This could be due to a memory access problem, or it could be you encountered a bug.

To help improve this program, please report bugs by e-mailing @.*** Please include details of how the problem you encountered can be reproduced_."

We checked the system memory, and it was sufficiently full to perform this tool. Additionally, we attempted to switch to another computer, but encountered the same error. here is our input file header : "SNP CHR BP A1 A2 BETA P SE FRQ N rs575272151 1 11008 G C -0.0303280157171869 0.0413221 0.014864608696787 0.0943995 247173 rs544419019 1 11012 G C -0.0303279257228457 0.0413229 0.0148646206629449 0.0943995 247173 rs62635286 1 13116 G T 0.0116022593832627 0.387424 0.0134238973205239 0.175915 247173 rs62028691 1 13118 G A 0.0116024132307965 0.387417 0.0134238786313126 0.175915 247173"

and

_"MARKER SNP WEIGHT N ALLELE A1 A2 EFFECT BETA STDERR SE FREQ FRQ PVAL P

PROCESS All_Breast_cancer.txt

SCHEME STDERR STDERR SE OVERLAP ON OUTFILE BC .TBL GENOMICCONTROL ON MINMAXFREQ ON AVERAGEFREQ ON ANALYZE Quit"_ looking forward to your reply###

This error only exist in window system, METAL perform well in linux

— Reply to this email directly, view it on GitHub https://github.com/statgen/METAL/issues/33#issuecomment-1593893149, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAIKQUC2NI7QLZ7LGBVXPQ3XLOUU3ANCNFSM6AAAAAAZH44ZSU . You are receiving this because you are subscribed to this thread.Message ID: @.***>

1667857557 commented 1 year ago

I think to solve the problem you probably need to download the Metal source and recompile. The old binaries can probably only address 2-4 GB of RAM. G … On Thu, Jun 15, 2023 at 8:50 PM Yu-Feng Huang @.> wrote: Dear researcher, Thank you for your work. Metal is a great tool in GWAS meta-analysis. However, when We including a summary GWAS file with over 800 million lines (an 846M file), it reports an error. "_ETAL HAS CRASHED The operating system has decided to terminate METAL. This could be due to a memory access problem, or it could be you encountered a bug. To help improve this program, please report bugs by e-mailing @. Please include details of how the problem you encountered can be reproduced." We checked the system memory, and it was sufficiently full to perform this tool. Additionally, we attempted to switch to another computer, but encountered the same error. here is our input file header : "SNP CHR BP A1 A2 BETA P SE FRQ N rs575272151 1 11008 G C -0.0303280157171869 0.0413221 0.014864608696787 0.0943995 247173 rs544419019 1 11012 G C -0.0303279257228457 0.0413229 0.0148646206629449 0.0943995 247173 rs62635286 1 13116 G T 0.0116022593832627 0.387424 0.0134238973205239 0.175915 247173 rs62028691 1 13118 G A 0.0116024132307965 0.387417 0.0134238786313126 0.175915 247173" and "MARKER SNP WEIGHT N ALLELE A1 A2 EFFECT BETA STDERR SE FREQ FRQ PVAL P PROCESS All_Breastcancer.txt SCHEME STDERR STDERR SE OVERLAP ON OUTFILE BC .TBL GENOMICCONTROL ON MINMAXFREQ ON AVERAGEFREQ ON ANALYZE Quit" looking forward to your reply### This error only exist in window system, METAL perform well in linux — Reply to this email directly, view it on GitHub <#33 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAIKQUC2NI7QLZ7LGBVXPQ3XLOUU3ANCNFSM6AAAAAAZH44ZSU . You are receiving this because you are subscribed to this thread.Message ID: @.***>

Jesus, I am completely unfamiliar with the C language compiler. Could you please provide me with more advice on how to solve this problem? For instance, I would greatly appreciate guidance on specific operations. Thank you very much.

Huang

AntiGravity commented 2 months ago

You can use the Linux version. The METAL crashes every time I perform it in RStudio, but it works well in Linux.