Excessive number of independent eQTLs

[zhoudreames]

I have 580 muscle samples of WGS and RNAseq data, and when using TensorQTL for cis-independent eQTL analysis, the number reached 170,000. My cis-independent eQTL TSS distribution is as follows, it seems that TSS enrichment is not concentrated enough. could you tell me the reason for the large number of identified samples?

This my code:

1.RNAseq mapping and Count quantification
hisat2 -x $ref -p 100 -I 0 --qc-filter -X 500 --dta -1 --rna-strandness RF ${ID}_RNAClean_1.fq.gz -2  ${ID}_RNAClean_2.fq.gz -S ${ID}.sam
 /home/guilu/software/star_2.7.9a_subread.simg featureCounts -T 40 -p -t exon -g gene_id -s 2 -Q 20 -C -a $GTF -o ${ID}.featureCounts.txt $BA

2.including gene filter with expression ≥0.1 TPM and ≥6 reads in ≥20% of samples. Then ,normalized the PCGs expression across samples within each tissue using the trimmed mean of M-value (TMM) method, implemented in edgeR4 (code is too long ,no show)

3. selceted 10 feer factor and 7 PCA factor as covariates

4.cis and indepent-cis eQTL analysis
singularity exec --nv ~/01.Biosoft/03.Sif/tensorqtl_GPU.simg python3  -m tensorqtl ${plink_prefix} ${expression_bed} ${name} \
--covariates ${covariates_file} \
--mode cis \
--seed 9823

singularity exec --nv ~/01.Biosoft/03.Sif/tensorqtl_GPU.simg python3 -m tensorqtl ${plink_prefix} ${expression_bed} ${name} \
--covariates ${covariates_file} \
--cis_output ${name}.cis_qtl.txt.gz \
--mode cis_independent

[francois-a]

Which file are you using to plot this? How does this distribution look for the eVariants from the initial --mode cis mapping?

[zhoudreames]

Which file are you using to plot this? How does this distribution look for the eVariants from the initial --mode cis mapping?

@francois-a Thank you for your reply. I used all.cis_independent_qtl.txt.gz to plot the above figure. the plot based on all.cis_qtl.txt.gz files files is as follows: