predictAge: Predict age using linear models

In zwdzwd/sesame: SEnsible Step-wise Analysis of DNA MEthylation BeadChips

Predict age using linear models

Description

The function takes a named numeric vector of beta values. The name attribute contains the probe ID (cg, ch or rs IDs). The function looks for overlapping probes and estimate age using different models.

Usage

predictAge(betas, model, na_fallback = FALSE, min_nonna = 10)

Arguments

betas	a probeID-named vector of beta values
model	a model object from sesameDataGet. should contain param, intercept, response2age. default to the Horvath353 model.
na_fallback	use fall back values if na
min_nonna	the minimum number of non-NA values.

Details

You can get the models such as the Horvath aging model (Horvath 2013 Genome Biology) from sesameDataGet. The function outputs a single numeric of age in years.

Here are some built-in age models: Anno/HM450/Clock_Horvath353.rds Anno/HM450/Clock_Hannum.rds Anno/HM450/Clock_SkinBlood.rds Anno/EPIC/Clock_PhenoAge.rds Anno/MM285/Clock_Zhou347.rds see vignette inferences.html#Age__Epigenetic_Clock for details

Value

age in the unit specified in the model (usually in year, but sometimes can be month, like in the mouse clocks).

Examples

betas <- sesameDataGet('HM450.1.TCGA.PAAD')$betas
## Not run: 
## download age models from
## https://github.com/zhou-lab/InfiniumAnnotationV1/tree/main/Anno
## e.g., Anno/HM450/Clock_Horvath353.rds
predictAge(betas, model)

## End(Not run)

zwdzwd/sesame documentation built on Dec. 21, 2024, 1:20 p.m.