Ana Cristina Gonzalez Sanchez
University of Bologna / Karolinska Institutet
Contact
GSA_Tissue_Celltype_Enrich is an R library intented to carry out in an easy manner Tissue enrichment analysis using data from GTEX, Celltype Enrichment Analysis using MAGMA-Celltyping or Geneset Enrichment Analysis for list of genes coming from differential expression analysis. Over representation of list of genes is tested in GWAS summary statistics.
if(!"devtools" %in% row.names(installed.packages())){
install.packages("devtools")
}
library(devtools)
if(!"GSA_Tissue_Celltype_Enrich" %in% row.names(installed.packages())){
install_github("Anacristina0914/GSA.Tissue.Celltype.Enrich")
}
library(GSA_Tissue_Celltype_Enrich) GWAS summary statistics must be formatted as specified in the MAGMA guidance. Briefly, the first three columns have to correspond to the SNP, CHR and BP (see example below):
| SNP | CHR | BP | P |
|---|---|---|---|
| rs1345 | 2 | 100123 | 0.01 |
| rs18667 | 3 | 30566921 | 0.5611 |
| rs145 | 16 | 9992021 | 0.173 |
This can be achieved either manually or using the MungeSumstats package described in Murphy & Skene 2021[1].
# Load required libraries for the package
load_required_libraries()
# Load GTEx gene expression data from Soma and Axon in human Motor Neurons. Only controls are loaded (Ctrl*).
Human_SomaAxon_data <- GSA.Tissue.Celltype.Enrich::load_GTEx_data_conditional(path = "/Soma_Axon_RNA-Seq/GSE121069_GEO_rpkms_human.txt",pattern = "Ctrl*",sep = "\t")
# Load GTEx data annotation
Human_SomaAxon_annot <- GSA.Tissue.Celltype.Enrich::load_GTEx_annot_conditional(path = "/Soma_Axon_RNA-Seq/", data = Human_SomaAxon_data, data_type = "Soma-Axon")
# Run Differential Expression analysis
tt_human_SomaAxon <- GSA.Tissue.Celltype.Enrich::run_diffExp_analysis(annot = Human_SomaAxon_annot, data = Human_SomaAxon_data, expr_path = "/Soma_Axon_RNA-Seq/", analysis_type = "D_Soma-Axon", species = "human")
# Map snps to genes and generate gene level p-value from summary statistics
genes.raw_path <- GSA.Tissue.Celltype.Enrich::map_snps_to_genes(gwas_path = "/ALS_sumstats.txt", N=NULL, genloc_filepath = "/genloc_files/NCBI37.3.gene.loc", genome_ref_path = "/g1000/g1000_eur",analysis_type = "D_Soma-Axon", species = "human")
# Run Gene Set Analysis (GSA) using MAGMA and up/down regulated genes from expression data and GWAS summary statistics.
GSA.Tissue.Celltype.Enrich::MAGMA_GSA(tt_filename = tt_human_SomaAxon, analysis_type = "D_Soma-Axon", genes.raw_path = genes.raw_path, species = "human", gene_n = 250)GTEx analysis v8 2017-06-05 gene-level median TPM by tissue dataset was retrieved from the GTEx database and further processed as described in Bryois et al. 2020[2] and its corresponding github repository jbryois/scRNA_disease. Briefly, bulk mRNA-seq data from 37 (14 brain and 27 non-brain) tissues was processed to obtain the specificity associated to each gene in each of the 37 tissues by dividing the expression of each gene in a tissue by the total expression of that gene in all tissues. Then, the 10% most specific genes per tissue were obtained(top10.txt) and used to test for enrichment in genetic associations in GWAS summary statistics.
For the GWAS summary statistics, variants with MinAlleleFreq < 0.01 were filtered out, as well as the Major Histocompatibility Complex (MHC) region (chr6: 25-34Mb) as described by Bryois et. at. 2020[2]. In order to accomplish this, the filter_Sumstats_MinAF.sh script was used.
exp_top10<- read_tsv("top10.txt")
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
[1]
Murphy, A. E., & Skene, N. G. (2021). MungeSumstats: A Bioconductor package for the standardisation and quality control of many GWAS summary statistics. bioRxiv.
[2]
Bryois, J., Skene, N. G., Hansen, T. F., Kogelman, L. J., Watson, H. J., Liu, Z., ... & Sullivan, P. F. (2020). Genetic identification of cell types underlying brain complex traits yields insights into the etiology of Parkinson’s disease. Nature genetics, 52(5), 482-493.