Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-2325
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-31T18:35:00.457434Z
status: published
lens: g_space
sector: health
headline: Bayesian Additive Regression Trees modeling of pharmacogenomic and clinical data predicts amlodipine response in hypertension
dek: The adoption of precision medicine in antihypertensive treatment remains challenging, largely because of substantial interindividual variability in drug response. This study aimed to determine whether genotyping combined with Bayesian machine-learning methods could improve the prediction of response to amlodipine. We conducted a multicenter prospective cohort study of 292 Chinese patients with hypertension treated with amlodipine. Eight single-nucleotide polymorphisms (SNPs) in key pharmacokinetic and pharmacody…
gain_title: A Bayesian machine-learning model combining six gene variants with clinical data predicted blood pressure control on amlodipine with an external validation AUC of 0.846, well above the comparison models.
problem_title: (none)
trace_subject: (none)
gain_reading: A Bayesian machine-learning model combining six gene variants with clinical data predicted blood pressure control on amlodipine with an external validation AUC of 0.846, well above the comparison models.
gain_evidence: The BART model achieved an external validation area under the receiver operating characteristic curve (AUC) value of 0.846, notably outperforming RFC (AUC = 0.61) and Bayesian linear models (AUC = 0.644-0.665).
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers in a multicenter prospective study followed 292 Chinese patients with hypertension who were treated with amlodipine. They sequenced eight gene variants involved in how the drug is processed and acts in the body, and combined six of them with clinical variables to build models predicting who would reach target blood pressure.

Doctors currently struggle to know in advance which patients will respond to a given blood pressure drug, so a better predictor could support more personalized prescribing. The findings come from one population and one drug taken alone, so it remains unclear how well the approach works in other groups or treatment regimens.
limitation: The cohort was limited to 292 Chinese patients on amlodipine monotherapy, and two of the eight gene variants had to be excluded because of insufficient variation, narrowing the genetic information used.
tag: Evidence-backed gain
key_points: The study enrolled 292 Chinese patients with hypertension treated with amlodipine in a multicenter prospective cohort. | Blood pressure control was defined as office systolic pressure under 140 mmHg and diastolic under 90 mmHg after treatment. | The Bayesian Additive Regression Trees model reached an external validation AUC of 0.846. | Random Forest Classification scored an AUC of 0.61, and Bayesian linear models scored 0.644 to 0.665. | Two of the eight gene variants were dropped because they showed too little variation.
rundown: The team tested eight single-nucleotide polymorphisms in pathways governing how amlodipine is absorbed, processed, and acts. At one locus, 217 individuals (99.5%) carried the wild-type genotype, and another was entirely homozygous for the mutant allele, so both were excluded from modeling.

The remaining six variants were merged with clinical variables to train prediction models. Bayesian Additive Regression Trees were compared with Random Forest Classification and Bayesian linear models using repeated cross-validation and external validation.

The authors conclude that combining pharmacogenomic data with advanced Bayesian machine learning offers a robust framework for personalized antihypertensive therapy, based on prediction of blood pressure control in patients on amlodipine monotherapy.
sources:
- peer_reviewed | Pharmacogenetics and Genomics | https://doi.org/10.1097/fpc.0000000000000617 | 2026-10-30
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

43f5c27c613a55d8ba9fafb027e8a1d5725765eb1f89c05d0fef5f3fbaaa7421

previous

0000000000000000000000000000000000000000000000000000000000000000