homeostaticdysregulation

homeostaticdysregulation

Mahalanobis distance from a young, healthy NHANES III reference cohort, fit separately by sex. The output is a log dysregulation score and is NOT expressed in years; larger values mean greater dysregulation. BioAge divides the score by the projection cohort's standard deviation of log(distance), so that NHANES IV constant is baked in per sex to keep single-sample predictions reproducible. C-reactive protein is supplied raw in mg/dL and log1p-transformed inside the clock. Sex is coded female = 1 and male = 0; a dataset with no female column scores every sample with the male reference. An incomplete panel biases the score downward: the score is a distance, so filling an absent biomarker with its reference value removes that marker's own contribution, which lowered the score for every one of the nine markers on average across the reference subjects and by as much as 2.9 on a 1.98-6.76 range. An incomplete panel therefore reads as healthier than it is, so the missing-feature warning the prediction pipeline emits should be heeded.

Model weights retain the original authors' terms; the pyaging software license does not relicense them.

Use with pyaging

import pyaging as pya

pya.pred.predict_age(adata, ["homeostaticdysregulation"])

Browse every clock in the pyaging Clock Catalogue.

Citation

Kwon, Dayoon, and Daniel W. Belsky. "A toolkit for quantification of biological age from blood chemistry and organ function test data: BioAge." GeroScience 43.6 (2021): 2795-2808.

https://doi.org/10.1007/s11357-021-00480-5