eQTL information for MR-JTI

[jcabanad]

Hi,

Thank you very much for sharing such an impressive work. I would like to run MR-JTI analysis but I don't know how to obtain data from eQTL, I mean beta, se and p-value. In addition, I thought that this will be a follow up control of pleiotropy after TWAS analysis, but following the tutorial I cannot see the connection between these steps. We just need to compare the results from both scripts?

Many thanks.

Best regards,

Judit

[jcabanad]

Regarding MR-JTI I have found that information in another question in this section:

MR-JTI uses SNP's marginal effect rather than the 'weights' from the prediction model. The marginal effect can be downloaded from the GTEx portal.

Could you please detail which file include this information in the GTEx portal?

Many thanks,

Judit

[zdangm]

Hi Judit, Sorry for the late reply. Just saw it. The MR-JTI takes the marginal effect size of eQTL, which is different from the "weight" in the prediction model, as input. If you have the access to the individual-level data, you could run a variant-gene cis eQTL association test to get the marginal effect size. The marginal effect size is also available from the GTEx portal. However, the platform started to charge for these large eQTL file downloads (via google cloud). As a potential compromise, you may use marginal effect size from the earlier GTEx version (v7) which is free to download. The estimate from v7 may be less accurate, but it should be unbiased. Dan

On Wed, Apr 28, 2021 at 3:39 PM jcabanad @.***> wrote:

Hi,

Thank you very much for sharing such an impressive work. I would like to run MR-JTI analysis but I don't know how to obtain data from eQTL, I mean beta, se and p-value. In addition, I thought that this will be a follow up control of pleiotropy after TWAS analysis, but following the tutorial I cannot see the connection between these steps. We just need to compare the results from both scripts?

Many thanks.

Best regards,

Judit

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/gamazonlab/MR-JTI/issues/5, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH5HAML6F7Q4RDWGGNVLFGDTLBXA7ANCNFSM43XZR2SA .

[zdangm]

Here is the link for v8. https://console.cloud.google.com/storage/browser/gtex-resources /gtex-resources/GTEx_Analysis_v8_QTLs/

v7 https://storage.googleapis.com/gtex_analysis_v7/single_tissue_eqtl_data/GTEx_Analysis_v7_eQTL_all_associations.tar.gz

Please let me know if you have any other questions. Dan

[jcabanad]

Thank you very much for you advice!

As you suggested, I'm using GTEx v7 data for my analysis, however I found an issue that I don't know how to manage. I would like to compute MR-JTI for those genes that overcome 5%FDR on the JTI analysis (as you did in your paper), about 35 genes in my case. However, I found that some SNPs are eQTL for more than one gene of interest in the GTEx data, so beta (slope) and standard_error (slope_se) differ among genes. As an example:

                 variant_id                        gene_id                           gene_id.1                  slope

174105 4_103374154_G_A_b37 ENSG00000164037 ENSG00000164037.12 -0.0611377 174106 4_103374154_G_A_b37 ENSG00000109332 ENSG00000109332.15 0.1121970 174107 4_103374154_G_A_b37 ENSG00000145354 ENSG00000145354.5 -0.0564166 slope_se pval_nominal alt rs_id_dbSNP147_GRCh37p13 174105 0.1332050 0.647127 A rs10000030 174106 0.0786011 0.156191 A rs10000030 174107 0.0876338 0.521013 A rs10000030

Should I run MR-JTI separately for each gene in those cases?

Thank you very much for your help.

Best regards,

Judit

[zdangm]

Hi Judit, Yes. You can run MR-JTI separately for each gene. MR-JTI handles the horizontal pleiotropy and provides a way to prioritize the causal gene in a region. The correlated (measured) expression could be the reason that you see multiple genes showing signals in a region. The correlated expression is a challenging thing for mapping the causal gene. Dan

[shishijingjing]

I would like to run MR-JTI analysis but I don't know how to obtain data ldsc.

[goodluckncb]

I am very surprised to find eqtl data storage can be downloaded here: https://yanglab.westlake.edu.cn/data/SMR/GTEx_V8_cis_eqtl_summary.html, and is free of charge