Get peak list

Detects chromatographic peaks, fits peak models, and extracts peak parameters at one or more wavelengths.

Usage

get_peaks(
  chrom_list,
  lambdas,
  fit = c("egh", "bemg", "gaussian", "raw"),
  sd_max = 50,
  max_iter = 100,
  time_unit = c("min", "s", "ms"),
  baseline = c("none", "flat", "sloped"),
  estimate_purity = FALSE,
  noise_threshold = 0.001,
  show_progress = NULL,
  cl = 2,
  collapse = FALSE,
  time.units = NULL,
  sd.max = NULL,
  max.iter = NULL,
  ...
)

Arguments

chrom_list: A list of chromatograms in matrix format (timepoints x wavelengths).
lambdas: A character or numeric vector specifying the wavelengths to find peaks at. For one-dimensional chromatograms, this argument can be ignored.
fit: What type of fit to use. Current options are bidirectional exponentially modified gaussian ("bemg"), exponential-gaussian hybrid ("egh"), "gaussian", or "raw". The raw setting performs trapezoidal integration directly on the raw data without fitting a model to the data. Defaults to egh.
sd_max: Maximum width (standard deviation) for peaks. Defaults to 50.
max_iter: Maximum number of iterations for non-linear least squares in fit_peaks.
time_unit: Specifies the units for sd, FWHM, area, and tau (if applicable). Options are minutes ("min"), seconds ("s"), or milliseconds ("ms"). Defaults to "min".
baseline: Whether to fit a baseline offset beneath each peak model. Options are "none" (no baseline, default), "flat" (constant offset per peak), or "sloped" (linearly varying offset per peak, useful for gradient elution). Not available when fit = "raw".
estimate_purity: Logical. Whether to estimate purity or not. Defaults to FALSE. (If TRUE, this will slow down the function significantly).
noise_threshold: Noise threshold. Argument to get_purity.
show_progress: Logical. Whether to show progress bar. Defaults to TRUE if pbapply is installed.
cl: Either an integer specifying the number of cores to use for parallel processing or a cluster object created by makeCluster. Defaults to 2. On Windows integer values will be ignored.
collapse: Logical. Whether to collapse multiple peak lists per sample into a single list when multiple wavelengths (lambdas) are provided.
time.units: The time.units argument is deprecated. Please use time_unit instead.
sd.max: The sd.max argument is deprecated. Please use sd_max instead.
max.iter: The max.iter argument is deprecated. Please use max_iter instead.
...: Additional arguments to find_peaks. Arguments provided to find_peaks can be used to fine-tune the peak-finding algorithm. Most importantly, the smooth_window should be increased if features are being split into multiple bins. Other arguments that can be used here include smooth_type, slope_thresh, and amp_thresh.

Value

The result is an S3 object of class peak_list, containing a nested list of data.frames containing information about the peaks fitted for each chromatogram at each of wavelengths specified by the lambdas argument. Each row in these data.frames is a peak and the columns contain information about various peak parameters:

rt: The retention time of the peak maximum.
start: The retention time where the peak is estimated to begin.
end: The retention time where the peak is estimated to end.
sd: The standard deviation (\(\sigma\)) of the fitted peak shape.
tau: Empirical skewness parameter (in units of time) determining peak asymmetry when fit == "egh".
tau_right: Exponential rate constant controlling right-sided tailing when fit = "bemg". Note that unlike tau in the EGH model, this parameter has units of 1/time.
tau_left: Exponential rate constant controlling left-sided tailing when fit = "bemg". Note that unlike tau in the EGH model, this parameter has units of 1/time.
h: The height parameter of the fitted BEMG shape function (only when fit = "bemg"). This is the value of the pure shape at center, before any baseline is added, and is distinct from height.
center: The center parameter of the fitted BEMG shape function (only when fit = "bemg"). This is the location around which the exponential tails are applied. Because BEMG peaks can be asymmetric (tau_right ≠ tau_left) and a sloped baseline shifts the fitted curve, the actual peak maximum (rt) may differ from center.
FWHM: The full-width at half maximum calculated as \(2.355 \sigma\).
height: The height of the peak above the fitted baseline.
area: The area of the peak above the fitted baseline, estimated by trapezoidal approximation.
r.squared: The coefficient of determination (\(R^2\)) of the fitted model to the raw data. (Note: this value is calculated by fitting a linear model of the fitted peak values to the raw data. This approach is statistically dubious, since the models are fit using non-linear least squares. Nevertheless, it can still be useful as a rough metric for "goodness-of-fit").
purity: The peak purity as estimated by get_purity.
floor: Constant baseline level (only when baseline = "flat").
floor_start: Baseline level at the left edge of the peak window (only when baseline = "sloped").
floor_end: Baseline level at the right edge of the peak window (only when baseline = "sloped"). Both floor_start and floor_end are constrained to be non-negative.

Details

Peaks are detected by finding zero-crossings in the smoothed first derivative of the specified chromatographic traces (using find_peaks). Peak models are then fit to the detected features using fit_peaks according to the value of fit. Available models include the bidirectional exponentially modified gaussian ("bemg"), exponential-gaussian hybrid ("egh") or "gaussian". Peak areas are then calculated using a trapezoidal approximation.

Additional arguments can be provided to find_peaks to fine-tune the peak-finding algorithm. For example, the smooth_window can be increased to reduce splitting of broad peaks into multiple features. Overly aggressive smoothing may cause small peaks to be overlooked.

The parameters sd, FWHM, tau, tau_right, tau_left, and area are returned in units determined by time_unit, which defaults to minutes. To compare directly with 'ChemStation' integration results, the time units should be changed to seconds.

Note

The bones of this function are adapted from the getAllPeaks function (https://github.com/rwehrens/alsace/blob/master/R/getAllPeaks.R) by Ron Wehrens (though the underlying algorithms for finding and integrating peaks are not the same.

References

Lan, K. & Jorgenson, J. W. 2001. A hybrid of exponential and gaussian functions as a simple model of asymmetric chromatographic peaks. Journal of Chromatography A 915:1-13. doi:10.1016/S0021-9673(01)00594-5 .
Naish, P. J. & Hartwell, S. 1988. Exponentially Modified Gaussian functions — A good model for chromatographic peaks in isocratic HPLC? Chromatographia, 26: 285-296. doi:10.1007/BF02268168 .
Arase, Shuntaro, Kanta Horie, Takashi Kato, et al. 2016. Intelligent Peak Deconvolution through In-Depth Study of the Data Matrix from Liquid Chromatography Coupled with a Photo-Diode Array Detector Applied to Pharmaceutical Analysis. Journal of Chromatography. A, 1469: 35–47. doi:doi.org/10.1016/j.chroma.2016.09.037 .
O'Haver, Tom. Pragmatic Introduction to Signal Processing: Applications in scientific measurement. https://terpconnect.umd.edu/~toh/spectrum/ (Accessed January, 2022).
Wehrens, R., Carvalho, E., Fraser, P.D. 2015. Metabolite profiling in LC–DAD using multivariate curve resolution: the alsace package for R. Metabolomics 11:143-154. doi:10.1007/s11306-014-0683-5 .

Author

Ethan Bass

Examples

if (FALSE) { # interactive()
data(Sa_pr)
pks <- get_peaks(Sa_pr, lambdas = c('210'), sd_max = 50, fit = "egh")
}