Evidence for Correlations Between BMI-Associated SNPs and circRNAs

The datasets provided here are part of the study "Evidence for Correlations Between BMI-Associated SNPs and circRNAs" by Rajcsanyi et al.

Abstract of the study:

Circular RNAs (circRNAs) are regulators of processes like adipogenesis. Their expression can be modulated by SNPs. We analysed links between BMI-associated SNPs and circRNAs. First, we detected an enrichment of BMI-associated SNPs on circRNA genomic loci in comparison to non-significant variants. Analysis of sex-stratified GWAS data revealed that circRNA genomic loci encompassed more genome-wide significant BMI-SNPs in females than in males. To explore if the enrichment is restricted to BMI, we investigated nine additional GWAS studies. We showed an enrichment of trait-associated SNPs in circRNAs for four analysed phenotypes (body height, chronic kidney disease, anorexia nervosa and autism spectrum disorder). To analyse the influence of BMI-affecting SNPs on circRNA levels in vitro, we examined rs4752856 located on hsa_circ_0022025. The analysis of heterozygous individuals revealed an increased level of circRNA derived from the BMI-increasing SNP allele. We conclude that genetic variation may affect the BMI partly through circRNAs.

Information regarding the datasets:

The data provided represents the analysed as well as generated data throughout the study. For further information about the datasets used, processed and generated, please see the study by Rajcsanyi et al.

circRNA datasets:

The analysed circRNA datasets were extracted from four circRNA databases (circAtlas v2.0, circBase, CIRCpediaV2 and circVAR) and were further processed to exclude internal duplicates and circRNAs derived from sex chromosomes. These original datasets have been downloaded from the following websites:

circAtlas v2.0: http://159.226.67.237:8080/new/links.php

circBase: http://www.circbase.org/cgi-bin/downloads.cgi

CIRCpediaV2: http://yang-laboratory.com/circpedia/download

circVAR: http://soft.bioinfo-minzhao.org/circvar/

GWAS datasets:

The original genome-wide association study (GWAS) summary statistics dataset of the BMI GWAS by Yengo et al. (2018) were classified into significant (P < 5*10-8) and non-significant (P >= 5*10-8) SNPs. A subsequent sensitivty analysis adjusted the P-value threshold of the non-significant SNPs to either P >= 5*10-5, P >= 5*10-6 or 5*10-7.
Please note that these datasets are not provided in this repository, as these were solely classified and divided based on the SNPs' P-value. The same applied for all additional GWAS data solely divided based on the P-value (GWAS data for Anorexia nervosa, Autism spectrum disorder, etc.). The original and complete summary statistcs data can be obtained in the stated references below for each GWAS. Yet, the datasets generated for an approximation of the linkage disequilibrium based on the GWAS data by Yengo et al. (2018, BMI) are provided in this repository.

BMI and body height: Yengo et al. (2018)

BMI sex-stratified: Pulit et al. (2019)

Anorexia nervosa: Watson et al. (2018)

Amyotropic lateral scerlosis: Iacoangeli et al. (2020)

Autism spectrum disorder: Grove et al. (2019)

Chronic kidney disease: Wuttke et al. 2019

Epilepsy: ILAE consortium et al. 2018

Heart Failure: Shah et al. 2020

Pernicious anemia: Glanville et al. 2021

Ulcerative Colitis: de Lange et al. 2017

Generated data:

The unprocessed output data of the study produced by the custom R script is provided here. Please note that the amount of information (rsID, P-value, Beta-value, allele frequency, circRNA_ID, circRNA strand, etc.) can vary between the output files due to differences in the data included in each circRNA and GWAS dataset. These dataset represent the raw and thus unprocessed output data. The results/counts presented in the study were obtained by further processing these output files. Further, the data produced by the SNaPshot assay are provided as well.