Estimate RT Drift — estimate_rt

Generate a mapping of observed sample retention times onto imputed (consistent) retention time.

estimate_rt_drift(
  sample_matched_compounds,
  previous_rt_update_key = NULL,
  sd_rt_resid_cutoff = 0.1,
  spline_ridge_penalty = 200,
  spline_degree = 4L,
  return_plot_data = FALSE
)

Arguments

sample_matched_compounds: tibble containing at least sampleId, compound_id and rt.
previous_rt_update_key: previous mapping of original RTs in raw files to RTs used in peaks and scans.
sd_rt_resid_cutoff: Cutoff for excluding a compound based on the standard deviation of its residuals (rt - fitted rt): sd(resid)/range(rt) < sd_rt_resid_cutoff.
spline_ridge_penalty: Ridge penalty on sample-level splines which model deviations between observed retention times and the consensus rt of a compound. Used for the H parameter in gam.
spline_degree: Degree of spline used to estimate drift (integer).
return_plot_data: return data which can be used to plot alignment summaries with ms2_driven_aligner_plotting

Value

a list containing sample specific gam fits and a sample independent fit which should be used to correct sample-level overfitting

Details

Alternate between estimating compound-specific retention time and sample-specific retention time deviations.