Identification of significant genes in genomics using Bayesian variable selection methods

Figures & data

Table 1 A panel of 42 SNPs by the CDC Chronic Fatigue Syndrome Research Group.

Gene	SNPs
POMC	rs12473543
TH	rs4074905, rs2070762
MAOA	rs1801291, rs979606, rs979605
MAOB	rs3027452, rs2283729, rs1799836
TPH2	rs2171363, rs4760816, rs4760750, rs1386486, rs1487280, rs1872824, rs10784941
COMT	rs4646312, rs740603, rs6269, rs4633, rs165722, rs933271, rs5993882
NR3C1	rs2918419, rs1866388, rs860458, rs852977, rs6196, rs6188, rs258750
SLC6A4	rs2066713, hCV7911132, rs140701
CRHR1	rs110402, rs1396862, rs242940, rs173365, rs242924, rs7209436
CRHR2	rs2267710, rs2267714, rs2284217

short-legendFigure 1

Table 2 In Stage I, the top ten SNPs based on their marginal posterior probabilities using the OpenBUGS software for three Bayesian variable selection methods including the Gibbs variable selection (GVS), the Stochastic Search Variable Selection (SSVS), the unconditional priors (UP) approach.

Posterior Ranking	GVS	SSVS	UP
1	hCV7911132	hCV7911132	hCV7911132
2	rs173365	rs1866388	rs173365
3	rs2070762	rs2070762	rs2070762
4	rs6196	rs173365	rs6196
5	rs1866388	rs1801291	rs1866388
6	rs140701	rs6196	rs2267710
7	rs2267710	rs2267710	rs140701
8	rs1801291	rs852977	rs1801291
9	rs852977	rs933271	rs852977
10	rs1396862	rs165722	rs1396862

Table 3 In Stage II, the top five SNPs based on their marginal posterior probabilities using the OpenBUGS software for three Bayesian variable selection methods including the Gibbs variable selection (GVS), the Stochastic Search Variable Selection (SSVS), the unconditional priors (UP) approach.

Posterior Ranking	GVS	SSVS	UP
1	hCV7911132	hCV7911132	hCV7911132
2	rs173365	rs2070762	rs173365
3	rs2070762	rs1866388	rs2070762
4	rs6196	rs173365	rs6196
5	rs1866388	rs6196	rs1866388