Skip to contents

glyclean

Omics data cleaning and preprocessing is a critical yet cumbersome step. glyclean helps you perform these tasks with ease, so you can focus on the fun part: downstream analysis!

Installation

You can install the latest release of glyclean from GitHub with:

# install.packages("remotes")
remotes::install_github("glycoverse/glyclean@*release")

Or install the development version:

remotes::install_github("glycoverse/glyclean")

Documentation

  • 🚀 Get started: Here
  • 📚 Reference: Here

Role in glycoverse

As data preprocessing is an essential step in omics data analysis, glyclean plays a central role in the glycoverse ecosystem. It serves as the bridge between raw experimental data (imported via glyread) and downstream analysis, enabling other packages like glystats to work with clean, analysis-ready data.

Example 

library(glyexp)
library(glyclean)
#> 
#> Attaching package: 'glyclean'
#> The following object is masked from 'package:stats':
#> 
#>     aggregate

exp <- real_experiment
clean_exp <- auto_clean(exp)
#> 
#> ── Normalizing data ──
#> 
#> No QC samples found. Using default normalization method based on experiment
#> type.
#> Experiment type is "glycoproteomics". Using `normalize_median()`.
#> 
#> 
#> ── Removing variables with too many missing values ──
#> 
#> 
#> 
#> No QC samples found. Using all samples.
#> Applying preset "discovery"...
#> Total removed: 24 (0.56%) variables.
#> 
#> 
#> ── Imputing missing values ──
#> 
#> 
#> 
#> No QC samples found. Using default imputation method based on sample size.
#> Sample size <= 30, using `impute_sample_min()`.
#> 
#> 
#> ── Aggregating data ──
#> 
#> 
#> 
#> Aggregating to "gfs" level
#> 
#> 
#> ── Normalizing data again ──
#> 
#> 
#> 
#> No QC samples found. Using default normalization method based on experiment
#> type.
#> Experiment type is "glycoproteomics". Using `normalize_median()`.
#> 
#> 
#> ── Correcting batch effects ──
#> 
#> 
#> 
#> ℹ Batch column  not found in sample_info. Skipping batch correction.

Yes, that’s it! Calling the magical auto_clean() function will automatically perform the following steps, in the most suitable way for your data:

  • Normalization
  • Missing value filtering
  • Imputation
  • Batch effect correction

and other steps that are necessary for downstream analysis.