Skip to contents

SEC/GPC Peak Deconvolution

Source: R/step_sec_peaks_deconvolve.R
step_sec_peaks_deconvolve.Rd

step_sec_peaks_deconvolve() creates a specification of a recipe step that resolves overlapping peaks in SEC/GPC chromatograms using curve fitting. This is a thin wrapper around measure::step_measure_peaks_deconvolve() with SEC-optimized defaults.

Usage

step_sec_peaks_deconvolve(
  recipe,
  model = "emg",
  optimizer = "auto",
  max_iter = 500L,
  quality_threshold = 0.9,
  smart_init = TRUE,
  constrain_positions = TRUE,
  peaks_col = ".peaks",
  measures_col = ".measures",
  role = NA,
  trained = FALSE,
  skip = FALSE,
  id = recipes::rand_id("sec_peaks_deconvolve")
)

Arguments

recipe

A recipe object. The step will be added to the sequence of operations for this recipe.

model

Peak model to use:

  • "emg" (default): Exponentially Modified Gaussian, recommended for SEC chromatograms with peak tailing

  • "gaussian": Symmetric Gaussian, for well-behaved symmetric peaks

  • "bigaussian": Bi-Gaussian for flexible asymmetry

EMG is the default because chromatographic peaks typically exhibit tailing due to mass transfer kinetics and column effects.

optimizer

Optimization method:

  • "auto" (default): Selects based on problem complexity

  • "lbfgsb": L-BFGS-B (fast, local optimization)

  • "multistart": Multiple starts for robustness

  • "nelder_mead": Derivative-free simplex method

max_iter

Maximum iterations for optimization. Default is 500.

quality_threshold

Minimum R-squared to accept fit. Default is 0.9 (stricter than base measure default of 0.8 for analytical quality).

smart_init

Logical. Use smart initialization based on peak properties. Default is TRUE. Highly recommended for SEC data.

constrain_positions

Logical. Enforce that peak centers maintain their relative ordering during optimization. Default is TRUE.

peaks_col

Name of the peaks column. Default is ".peaks".

measures_col

Name of the measures column containing the chromatogram. Default is ".measures" but often "ri", "uv", etc. in SEC workflows.

role

Not used by this step.

trained

A logical to indicate if the step has been trained.

skip

A logical. Should the step be skipped when baking?

id

A character string that is unique to this step.

Value

An updated version of recipe with the new step added. The .peaks column will be updated with:

  • Refined peak parameters from curve fitting

  • fit_r_squared: R-squared of the overall fit

  • area: Integrated area under the fitted curve (analytical integration)

Details

Peak deconvolution is essential for SEC/GPC analysis when peaks overlap, which is common for:

  • Bimodal molecular weight distributions

  • Aggregate/monomer/fragment species in protein SEC

  • Copolymer component separation

EMG Model (Default):

The Exponentially Modified Gaussian (EMG) is ideal for SEC because it accounts for the characteristic tailing observed in chromatographic peaks:

$$EMG(x) = h \cdot \exp(\sigma^2/(2\tau^2) + (c-x)/\tau) \cdot \Phi((x-c)/\sigma - \sigma/\tau)$$

where \(h\) is height, \(c\) is center, \(\sigma\) is Gaussian width, and \(\tau\) is the exponential decay parameter (tailing).

Quality Assessment:

The quality_threshold parameter sets the minimum acceptable R-squared for fits. The default of 0.9 is stricter than the base measure default, appropriate for quantitative analytical work.

Tidying

When you tidy() this step, a tibble with columns model, optimizer, quality_threshold, and id is returned.

See also

step_sec_peaks_detect() for SEC-optimized peak detection, measure::step_measure_peaks_deconvolve() for the underlying implementation.

Other sec-chromatography: step_sec_band_broadening(), step_sec_baseline(), step_sec_detector_delay(), step_sec_exclude_regions(), step_sec_integration_window(), step_sec_mw_averages(), step_sec_mw_distribution(), step_sec_mw_fractions(), step_sec_peaks_detect(), step_sec_peaks_refine()

Examples

if (FALSE) { # \dontrun{
library(recipes)
library(measure)

# SEC peak deconvolution with EMG model (default)
rec <- recipe(~., data = sec_data) |>
  step_measure_input_long(ri_signal, location = vars(elution_volume)) |>
  step_sec_peaks_detect() |>
  step_sec_peaks_deconvolve() |>
  prep()

# Use Gaussian model for symmetric peaks
rec <- recipe(~., data = sec_data) |>
  step_measure_input_long(ri_signal, location = vars(elution_volume)) |>
  step_sec_peaks_detect() |>
  step_sec_peaks_deconvolve(model = "gaussian") |>
  prep()

# Use multistart optimizer for complex multi-peak scenarios
rec <- recipe(~., data = sec_data) |>
  step_measure_input_long(ri_signal, location = vars(elution_volume)) |>
  step_sec_peaks_detect() |>
  step_sec_peaks_deconvolve(optimizer = "multistart") |>
  prep()
} # }