Identifying causal genes for stroke via integrating the proteome and transcriptome from brain and blood

Table 3 Candidate genes identified by Mendelian randomization, Bayesian colocalization and Steiger filtering analysis in brain

Data source	Stroke subytpe	Gene	Beta	SE	P-value	PPH4	Correct Direction	Steiger_P Value	If replicated in PWAS/TWAS
pQTL	AS	ALDH2	0.639	0.150	2.06E−05	60.20%	TRUE	8.78E−14	Yes
	AIS	ALDH2	0.748	0.164	5.00E−06	17.30%	TRUE	9.43E−14	Yes
	LAS	–
	CES	–
	SVS	ICA1L*	− 2.521	0.569	9.49E−06	99.20%	TRUE	1.22E−13	Yes
eQTL	AS	HDAC9	0.253	0.041	5.19E−10	100.00%	TRUE	1.21E−09	No
	AIS	HDAC9	0.276	0.044	3.38E−10	100.00%	TRUE	1.24E−09	No
		HECTD4	0.242	0.050	1.44E−06	95.10%	TRUE	4.03E−09	No
	LAS	HDAC9	0.757	0.106	1.06E−12	100.00%	TRUE	1.59E−09	No
	CES	–
	SVS	ICA1L*	0.229	0.046	5.73E−07	87.20%	TRUE	3.89E−36	Yes
		CARF	0.280	0.057	7.14E−07	81.60%	TRUE	3.79E−23	No
		ADRB1	− 0.359	0.075	1.47E−06	91.10%	TRUE	1.52E−33	No
		NBEAL1	0.317	0.068	2.64E-06	79.40%	TRUE	7.87E−17	Yes

Table shows the beta, SE and P values for the MR analysis of brain pQTL (above) and eQTL (down). PPH4 denotes the posterior probability that two traits share a causal genetic variant using Bayesian colocalization analysis. Correct Direction and P value are given for Steiger filtering analysis, which shows the correct direction for the effect between exposure and stroke risk in this table. * indicated the gene which was identified in the MR analysis using both pQTL and eQTL data

ISSN: 1479-5876