GTEx TPMs Vs normalized data

Question

[DS]

Hi all,

I am using the GTEx cohort, and based on the published datasets, I would like to choose specific genes that are not included in the GTEx_Analysis_v8_eQTL_expression_matrices.tar (which have been completely processed, filtered, and normalized).

Can someone please provide clarification on the necessary steps for me to follow in order to ensure that if I extract the genes of interest from the TPM matrices GTEx_Analysis_2017-06-05_v8_RNASeQCv1.1.9_gene_tpm.gct.gz my gene filters and normalisation will be directly comparable?

Thanks!

[jv]

my gene filters and normalisation will be directly comparable?

Directly comparable to what? What analysis do you want to perform?

[DS]

I would like to conduct eQTL analysis on certain genes that are not included in the GTEx_Analysis_v8_eQTL_expression_matrices.tar. Therefore, I need to extract these genes from the GTEx_Analysis_2017-06-05_v8_RNASeQCv1.1.9_gene_tpm.gct.gz. I am not sure about the steps I need to take to process/normalize the TPMs.

I am sorry if this is a very basic question, but I cannot find anything on the GTEx portal.

Answer 1

[jv]

It looks like the gtex-pipeline repo has info on how the files in GTEx_Analysis_v8_eQTL_expression_matrices.tar were prepared:

eqtl_prepare_expression.py ${tpm_gct} ${counts_gct} ${annotation_gtf} \
    ${sample_participant_lookup} ${vcf_chr_list} ${prefix} \
    --tpm_threshold 0.1 \
    --count_threshold 6 \
    --sample_frac_threshold 0.2 \
    --normalization_method tmm

https://github.com/broadinstitute/gtex-pipeline/blob/master/qtl/eqtl_prepare_expression.wdl

https://github.com/broadinstitute/gtex-pipeline/tree/master/qtl