AUTHOR
LOIC YENGO,
Institute for Molecular Bioscience,
The University of Queensland;
Contact: l.yengo@imb.uq.edu.au


METHODS

GWAS analysis: ANA_C2_V2
Download: https://storage.googleapis.com/covid19-hg-public/20200619/results/build_37/COVID19_HGI_ANA_C2_V2_20200629.txt.gz.b37_1.0E-5.txt

eQTL data: eQTLgen (link to follow)
Tissue: whole blood
Reference:

mQTL data: 
Download: https://www.dropbox.com/s/os4cgkb4519wbvn/US_mQTLS_SMR_format.zip?dl=0
Tissue: Whole blood
Reference: Hannon et al. AJHG (2017): https://www.sciencedirect.com/science/article/pii/S0002929717301581?via%3Dihub

SMR analyses
Version: 1.03
Command line:

/home/l.yengo/exe/smr_Linux \
	--bfile geno/chrom${chrom} \
	--gwas-summary ../download/29JUN2020/COVID19_HGI_ANA_C2_V2_20200629.txt.gz.b37_1.0E-5.ma \
	--beqtl-summary ${prefix} --cis-wind 2000 --thread-num 10 \
	--out results/${QTLtype}/chrom${chrom}.${code}.${tissue}

Genotype: randomly selected 10,000 participants of the UK Biobank. Imputed data (HRC+1000 genomes). Hard-called genotypes (posterior probability >0.9 otherwise set to missing). Missing value threshold 5%. Hardy-Weinberg Equilibrium p-value threshold: 1e-5. Minor allele count >5.


RESULTS

eQTL results 
-	One probe (ENSG00000183625) for the C-C chemokine receptor type 3 (CCR3) is detected at p_SMR<5e-8 (and p_HEIDI>0.05). HEIDI test checks if colocalization is due to causal variants being in LD as opposed to pleiotropy or causality.
-	Effect size: increased expression of CCR3 correlates positively with increased susceptibility to COVID-19. 
-	eQTL position (hg19): 3:45909024
-	cis-Heritability of CCR3 expression: 0.72 (S.E. 0.03).


mQTL results 
-	Two probes (cg00634029 and cg02315645) near LZTFL1 and CCR2 respectively detected at p_SMR<5e-8. Not enough SNPs for HEIDI test.
-	Effect size: increased methylation at these sites correlates positively with increased susceptibility to COVID-19.
-	mQTL position (hg19): cg00634029 (3:45889949) and cg02315645 (3:45908514)