********************************************************* * GCTB 2.5.6 beta * * Genome-wide Complex Trait Bayesian analysis * * For inquiries, contact: Jian Zeng * * Last updated: 24 May, 2026 * * MIT License * ********************************************************* Options: --ldm-eigen ldm_eigen --gwas-summary gwas.imputed.ma --annot annotations.txt --sbayes RC --chain-length 1000 --burn-in 400 --thread 4 --out sbayesrc Using 4 threads. Analysis started: Sat Jun 6 22:10:04 2026 Reading LD matrix eigen-decomposition data... Reading LDM info from file [ldm_eigen/ldm.info]. 590 LD Blocks to be included from [ldm_eigen/ldm.info]. Reading LDM SNP info from file [ldm_eigen/snp.info]. 272826 SNPs to be included from [ldm_eigen/snp.info]. Reading SNP annotation from [annotations.txt]. 272826 matched SNPs in the annotation file (12 annotations and 136754 SNPs have more than one annotation). Reading GWAS summary data from [gwas.imputed.ma]. 272826 matched SNPs in the GWAS summary data (in total 272826 SNPs). 272826 SNPs on 22 chromosomes are included. 590 LD blocks are included. GWAS effect scaling: estimated by (1/sqrt(n*SE^2+b^2) from summary stats. Reading eigenvectors from binary file and making W and Q matrices... Data summary: mean sd GWAS SNP Phenotypic variance 210.528 12.583 GWAS SNP heterozygosity 0.353 0.126 GWAS SNP sample size 2997 3 GWAS SNP effect (in genotype SD unit) -0.000 0.018 GWAS SNP SE 0.018 0.000 LD block size 462 163 LD block rank 328 96 Annotation info: 1 Intercept 272826 1.000 2 Anno1 99 0.000 3 Anno2 99 0.000 4 Anno3 100 0.000 5 Anno4 100 0.000 6 Anno5 99 0.000 7 Anno6 99 0.000 8 Anno7 99 0.000 9 Anno8 99 0.000 10 Anno9 99 0.000 11 Anno10 100 0.000 12 Anno11 136481 0.500 Finding the best eigen cutoff from [0.995 0.99 0.95 0.9] based on pseudo summary data validation. Cutoff Prediction accuracy (r) Relative accuracy 0.995 0.556542 1 0.99 0.58244 1.04653 0.95 0.562921 1.01146 0.9 0.525175 0.943639 0.995 is selected to be the eigen cutoff to continue the analysis (time used: 0:0:33). SBayesRC Using the low-rank model scale factor: 0.000807344 Gamma: 0 0.001 0.01 0.1 1 Fitting model assuming scaled genotypes MCMC launched ... Number of chains: 1 Chain length: 1000 iterations Burn-in: 400 iterations Iter NumSnp1 NumSnp2 NumSnp3 NumSnp4 NumSnp5 Vg1 Vg2 Vg3 Vg4 Vg5 SigmaSq hsq ResVar NumSkeptSnp TimeLeft 100 270133 1637 901 87 67 0.0000 0.0158 0.0769 0.0915 0.8158 0.0065 0.5880 1.0099 0 0:3:0 200 270241 1979 496 27 82 0.0000 0.0166 0.0463 0.0276 0.9095 0.0063 0.5463 1.0080 0 0:2:40 300 270270 1836 599 15 105 0.0000 0.0152 0.0459 0.0112 0.9276 0.0075 0.6937 1.0098 1 0:2:22 400 270404 1747 584 12 77 0.0000 0.0149 0.0490 0.0107 0.9253 0.0052 0.5431 1.0090 0 0:2:4 500 270516 1933 289 8 78 0.0000 0.0170 0.0282 0.0031 0.9517 0.0055 0.5133 1.0123 0 0:1:45 600 270281 2364 79 2 98 0.0000 0.0223 0.0066 0.0023 0.9688 0.0051 0.5083 1.0121 0 0:1:24 700 269940 2763 25 0 96 0.0000 0.0286 0.0028 0.0000 0.9686 0.0053 0.5065 1.0145 0 0:1:3 800 270285 2426 11 0 102 0.0000 0.0209 0.0006 0.0000 0.9785 0.0053 0.5612 1.0100 0 0:0:42 900 270215 2496 11 4 98 0.0000 0.0193 0.0013 0.0014 0.9780 0.0051 0.5418 1.0117 0 0:0:21 1000 270093 2633 3 15 80 0.0000 0.0241 0.0001 0.0134 0.9624 0.0047 0.4695 1.0096 0 0:0:0 MCMC cycles completed. Posterior statistics from MCMC samples: Parameter Mean SD NumSnp1 270251.718750 156.767349 NumSnp2 2355.716309 290.854767 NumSnp3 122.250015 181.753098 NumSnp4 3.233334 3.913084 NumSnp5 91.116669 8.691265 Vg1 0.000000 0.000000 Vg2 0.021375 0.003100 Vg3 0.011131 0.017129 Vg4 0.002494 0.004162 Vg5 0.965000 0.016911 SigmaSq 0.005276 0.000320 hsq 0.525862 0.029128 ResVar 1.010328 0.002652 Ouput 2 skeptical SNPs in [sbayesrc.skepticalSNPs], whose posterior joint effect sizes are remarkably greater than their marginal effect sizes. Since this may be due to poor convergence, their posterior effects will be set to be zero. Analysis finished: Sat Jun 6 22:14:20 2026 Computational time: 0:4:16