Chromatogram Artifact Removal

chromatogram-artifact-removal

Summary

Remove instrumental noise and baseline drift from 2D-TIC (two-dimensional Total Intensity Chromatogram) GCxGC-MS data using asymmetric least squares baseline correction and Whittaker smoothing to enhance signal quality and reveal metabolite differences. This preprocessing step is essential before peak alignment and multivariate analysis.

When to use

Raw NetCDF-format GCxGC-MS chromatograms exhibit steady or increasing baseline intensity caused by instrumental contamination, column bleeding, or thermal drift; smoothing and baseline correction are needed before peak alignment (2DCOW) to avoid alignment artifacts and to improve the signal-to-noise ratio for downstream multivariate analysis (MPCA).

When NOT to use

• Input is already baseline-corrected or has been manually pre-processed by external software — applying baseline_corr redundantly may distort true metabolite signals.
• The 2D chromatogram is from a non-GCxGC instrument or is stored in a format other than NetCDF (e.g., vendor-specific binary formats not imported via read_chrom).
• Analysis goal is only qualitative identification of major peaks; baseline correction introduces algorithmic assumptions that may not be justified for single-sample exploratory work.

Inputs

• 2D-TIC chromatogram object (output from read_chrom function, derived from NetCDF file)

Outputs

• Baseline-corrected and smoothed 2D-TIC chromatogram object
• Preprocessed chromatogram ready for peak alignment and multivariate analysis

How to apply

After importing and folding the raw chromatogram into 2D-TIC using read_chrom, apply smoothing first using the wsmooth function with the Whittaker smoother algorithm to attenuate high-frequency chemical and instrumental noise. Then apply baseline_corr using the asymmetric least squares algorithm to remove the slowly-varying baseline component. The asymmetric least squares method uses differential penalties for positive and negative residuals, allowing it to preserve sharp metabolite peaks while suppressing baseline undulation. Order matters: smoothing before baseline correction prevents the correction step from over-fitting to noise. Visually inspect the corrected 2D-TIC to confirm that baseline distortion is removed without peak attenuation and that the separation space now contains more intense, well-resolved signals.

Related tools

• RGCxGC (Provides baseline_corr function (asymmetric least squares baseline correction) and wsmooth function (Whittaker smoothing) for 2D-TIC artifact removal) — https://github.com/DanielQuiroz97/RGCxGC
• R (Execution environment for RGCxGC package functions)

Examples

baseline_corr(chrom_obj, lambda=100, p=0.001); wsmooth(chrom_obj, lambda=100)

Evaluation signals

• Visual inspection of the baseline-corrected 2D chromatogram shows a flat, near-zero baseline with no increasing or oscillating drift.
• Peak intensities in the separation space are increased relative to the baseline (higher signal-to-noise ratio) compared to the raw chromatogram.
• Peaks remain sharp and well-resolved; no significant broadening or distortion is visible after smoothing and baseline correction.
• Downstream peak alignment (2DCOW) produces fewer alignment artifacts and higher alignment scores when applied to preprocessed chromatograms versus raw data.
• Multivariate analysis (MPCA) on corrected chromatograms reveals clearer group separation and higher explained variance in the first few components compared to unprocessed data.

Limitations

• Asymmetric least squares baseline correction assumes baseline distortion is smooth and slowly-varying; sharp instrumental spikes or hot-spot contamination may not be fully removed.
• The choice of smoothing strength (lambda parameter in Whittaker smoother) and baseline correction asymmetry (p parameter in asymmetric least squares) can impact peak shape and quantification; no universal defaults are described in the article.
• Over-smoothing can eliminate low-intensity or poorly-resolved metabolite peaks; under-smoothing leaves residual noise that degrades alignment and multivariate analysis.
• Baseline correction is data-dependent and may not perform equally well across all retention time regions if baseline behavior varies locally.

Evidence

• [other] baseline_corr function using asymmetric least squares remove baseline distortion from a 2D-TIC chromatogram: "The baseline_corr function applies the asymmetric least squares algorithm to remove steady and increasing intensity from 2D-TIC chromatograms caused by instrumental contamination or column bleeding"
• [intro] RGCxGC offers preprocessing algorithms including baseline correction and smoothing: "RGCxGC offers common pre-processing algorithms for signal enhancement, such as baseline correction based on asymetric least squares, smoothing based on the Whittaker smoother"
• [intro] Preprocessing steps reduce noise and enable multivariate analysis: "noise can be significantly reduced by using preprocessing algorithms, herein smoothing, baseline correction, and peak alignment. Then, in order to reveal differences between groups, multivariate"
• [intro] workflow order: first read_chrom, then apply wsmooth and baseline_corr: "you can perform preprocessing routines like smoothing and/or baseline correction by using the function wsmooth and baseline_corr"
• [readme] Raw chromatographic signals contain undesirable artifacts requiring preprocessing: "Raw chromatographic signals usually contains undesirable artifacts, such as chemical and"