Step: Cox Proportional Hazards Model

Perform survival analysis by fitting a Cox proportional hazards model using glystats::gly_cox() for each variable. This step identifies variables associated with survival outcomes.

This step depends on the on parameter (default: exp).

• When on = "exp", requires exp (usually after step_preprocess()).

• When on = "sig_exp", requires sig_exp from one of step_dea_limma(), step_dea_ttest(), step_dea_wilcox(), step_dea_anova(), or step_dea_kruskal().

• When on = "trait_exp", requires trait_exp from step_derive_traits().

• When on = "sig_trait_exp", requires sig_trait_exp from DEA on traits.

• When on = "dynamic_motif_exp", requires dynamic_motif_exp from step_quantify_dynamic_motifs().

• When on = "sig_dynamic_motif_exp", requires sig_dynamic_motif_exp from DEA on motifs.

• When on = "branch_motif_exp", requires branch_motif_exp from step_quantify_branch_motifs().

• When on = "sig_branch_motif_exp", requires sig_branch_motif_exp from DEA on motifs.

Usage

step_cox(
  on = "exp",
  time_col = "time",
  event_col = "event",
  p_adj_method = "BH",
  ...
)

Arguments

on

Name of the experiment to run Cox regression on. Can be "exp", "sig_exp", "trait_exp", "sig_trait_exp", "dynamic_motif_exp", "sig_dynamic_motif_exp", "branch_motif_exp", "sig_branch_motif_exp".

time_col

Column name in sample information containing survival time. Default is "time".

event_col

Column name in sample information containing event indicator (1 for event, 0 for censoring). Default is "event".

p_adj_method

Method for adjusting p-values. See p.adjust.methods. Default is "BH". If NULL, no adjustment is performed.

...

Additional arguments passed to glystats::gly_cox().

Value

A glysmith_step object.

Details

Data required:

• exp (if on = "exp"): The experiment to run Cox regression on

• trait_exp (if on = "trait_exp"): The trait experiment to run Cox regression on

• dynamic_motif_exp (if on = "dynamic_motif_exp"): The dynamic motif experiment to run Cox regression on

• branch_motif_exp (if on = "branch_motif_exp"): The branch motif experiment to run Cox regression on

The experiment must contain survival data with time_col and event_col columns in the sample information.

Tables generated (with suffixes):

• cox: A table containing Cox regression results with columns:

  • variable: Variable name

  • coefficient: Regression coefficient (log hazard ratio)

  • std.error: Standard error of the coefficient

  • statistic: Wald test statistic

  • p_val: Raw p-value from Wald test

  • hr: Hazard ratio (exp(coefficient))

  • p_adj: Adjusted p-value (if p_adj_method is not NULL)

See also

glystats::gly_cox(), survival::coxph()

Examples

step_cox()
#> <step "step_cox()"> Cox proportional hazards model
step_cox(time_col = "survival_time", event_col = "death")
#> <step "step_cox(time_col = \"survival_time\", event_col = \"death\")"> Cox
#> proportional hazards model
step_cox(on = "sig_exp", p_adj_method = "bonferroni")
#> <step "step_cox(on = \"sig_exp\", p_adj_method = \"bonferroni\")"> Cox
#> proportional hazards model of significant variables