No overlapping GWAS Z-score

[corunesa]

Hello!
When I do the TWAS with the weights, my summary and LDREF I have no problem. When I add --coloc_P, it tells me that there is no overlapping GWAS Z-scores. I can't find the bug and I don't understand why there is no problem when I don't have the colocalization flag.
Thank you very much!

[corunesa]

The title is wrong, the no overlapping problem is with GWAS Z-score

[sashagusev]

Hi corunesa, can you paste in the full FUSION command you are running with and without the coloc flag? Do you have all the parameters specified as described here:

http://gusevlab.org/projects/fusion/#colocalization-analysis-with-coloc

[corunesa]

Thank you very much for your response.
This is the command I use for coloc:

declare trait=“trait"

declare programa="/media/stroke/Almacen/NAS/GENETICA/programas/FUSION/"

declare weights="/media/stroke/Almacen/NAS/GENETICA/Datos/eQTL/"

declare ldref="/media/stroke/Almacen/NAS/GENETICA/Datos/1000G/LDREF/"

declare analisis="/media/stroke/Almacen/NAS/GENETICA/analisis/"

for tissue in Brain_Frontal_Cortex_BA9; do for chr in 2; do Rscript $programa/FUSION.assoc_test.R \

--sumstats $analisis/TWAS/$trait/hg19/tissues/snp_union_imputed.${trait}.gz \

--weights $weights/GTExv8.EUR.${tissue}.pos \

--weights_dir $weights \

--ref_ld_chr $ldref/1000G.EUR. \

--chr $chr \

--coloc_P 0.05 \

--GWASN 400417 \

--out $analisis/TWAS/$trait/hg19/tissues/$tissue/coloc/coloc_${trait}_${tissue}hg19${chr}.dat; done; done

This is the command for the TWAS (without problem):

declare data="/media/stroke/Almacen/NAS/GENETICA/Datos/SUMMARIES/6.IMPUTED_SUMMARIES/"
declare trait=“trait"
declare programa="/media/stroke/Almacen/NAS/GENETICA/programas/FUSION/"
declare weights="/media/stroke/Almacen/NAS/GENETICA/Datos/eQTL/"
declare ldref="/media/stroke/Almacen/NAS/GENETICA/Datos/1000G/LDREF/"
declare analisis="/media/stroke/Almacen/NAS/GENETICA/analisis/"

for tissue in Brain_Amygdala; do for a in $(seq 1 22); do Rscript $programa/FUSION.assoc_test.R
--sumstats $data/$trait/union_imputed.${trait}.gz
--weights $weights/GTExv8.EUR.${tissue}.pos
--weights_dir $weights/
--ref_ld_chr $ldref/1000G.EUR.
--chr ${a}
--out $analisis/TWAS/$trait/hg19/tissues/${tissue}/${trait}_hg19_chr${a}GTEx8${tissue}.dat; done; done

THANK YOU VERY MUCH AGAIN!

[sashagusev]

Sorry if I'm misinterpreting, but aren't these two runs using a different GWAS file: --sumstats analisis/TWAS/{trait}.gz versus --sumstats data/${trait}.gz ? Does the first command still break if you use --sumstats data/${trait}.gz?