In Silico Research in Biomedicine · Nov 2026
Blood cis-eQTL Mendelian Randomisation and Colocalisation among Candidate Targets of a Huangqi Eight-Herb Formula: TP53 Is the Only Supported Candidate
By Hu Y
Background Buzhong Yiqi Decoction, a Huangqi-led eight-herb formula, treats Wei syndrome, the Chinese-medicine counterpart of sarcopenia. Whether any candidate target is causally linked to muscle mass or strength is unknown. Methods Constituents were screened in TCMSP and targets constrained in four independent skeletal-muscle RNA-seq cohorts (179 individuals), giving nine core targets by enrichment and STRING topology. Docking, 100-ns molecular dynamics and MM-GBSA used metformin as a negative control. Every target with cis-eQTL data underwent two-sample Mendelian randomisation in whole blood and Bayesian colocalisation; colocalisation was the primary criterion and the FDR screen is exploratory. Results Screening yielded 182 compounds, 2 797 herb-compound-target relations, 25 intersecting and nine core targets. Of 112 tests, 17 reached FDR < 0.05, but only TP53 colocalised: strongly with appendicular lean mass and fat-free mass (PPH4 = 0.996 both) and weakly with right-hand grip strength (0.806). Higher genetically predicted TP53 expression was associated with lower muscle mass. In the largest skeletal-muscle eQTL resource (n = 816) no TP53 variant survived correction, and of the nine core targets only CASP3 could be instrumented, with a null estimate; the evidence is confined to blood cis-eQTL and is not muscle-specific. Docking did not establish constituent regulation of TP53: formononetin and kaempferol bound only weakly (-6.6 and -6.8 kcal/mol). Exploratory machine-learning, deconvolution and virtual-knockout modules were negative. Conclusions Among the candidate targets evaluated, TP53 alone is supported by blood cis-eQTL randomisation together with colocalisation. This genetic support does not show that any constituent of the formula regulates TP53.