Beyond Prior Weights: Receptor Contribution Reallocation and Hypothesis-Generating TWAS Candidates in Antipsychotic Metabolic-Risk Modelling

BackgroundAntipsychotic-associated metabolic abnormalities, including weight gain, dyslipidemia, hyperglycemia, obesity, and metabolic syndrome, differ substantially among drugs. Receptor pharmacology has provided important mechanistic clues, whereas transcriptome-wide association studies have identified genetically regulated expression signals associated with lipid traits. These evidence streams are usually evaluated separately.ObjectiveThis study performed a mechanistic downstream analysis of an integrated Ki–defined daily dose–transcriptome-wide association study pipeline. The analysis compared a V4 negative-log10-p-value scaling model using literature-derived metabolic receptor weights with an otherwise identical model in which all recognized receptor genes received a uniform weight of 1.0.MethodsOnly previously generated pipeline outputs were used. Long-format receptor-level files and per-drug risk-score files were analyzed for 52 antipsychotics across low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, log-transformed triglycerides, non-high-density lipoprotein cholesterol, and total cholesterol. Analyses included reconstruction validation, pooled receptor contribution, mapping of TWAS signal against literature weight, clinical-liability discordance, leave-one-gene-out sensitivity, TWAS-scaling sensitivity, and receptor co-occurrence in high-risk drug sets.ResultsReconstruction reproduced the saved risk ranks with a Spearman correlation of 1.000 for all ten mode–trait combinations. The V4 rankings remained highly concordant after removal of literature weights, with trait-specific Spearman correlations ranging from 0.900 to 0.923 and a mean of 0.913. The weighted model was dominated by HRH1, HTR2A, and HTR2C, which contributed 25.7%, 21.6%, and 10.3% of pooled signal, respectively. Under uniform weighting, DRD2, DRD3, and HTR2A became the leading contributors, accounting for 22.1%, 12.9%, and 12.6%. ADRB1 was the strongest low-weight TWAS candidate, with an absolute z score of 13.53 for HDL and a literature weight of 0.08. Other candidates included CHRM4, DRD2, DRD4, SLC6A4, and ADRB2. The model most strongly under-ranked paliperidone relative to the supplied clinical comparator, while ziprasidone rose substantially under uniform weighting. Removing HRH1 produced the largest rank shifts, particularly for levomepromazine, promazine, and acepromazine.ConclusionsRemoval of literature-derived weights did not collapse the overall drug ranking, but it substantially changed the receptor-level explanation of the score. The weighted model emphasized established histaminergic and serotonergic metabolic mechanisms, whereas the uniform model revealed broader dopaminergic and TWAS-associated contributions. ADRB1 should be considered a high-priority hypothesis-generating candidate rather than a confirmed causal mechanism. The framework supports dual reporting of a clinically anchored weighted model and a discovery-oriented uniform model.