A Transparent Analysis Pipeline for Mammalian Metabolic Phenotype Microarray Data

Provides a transparent, end-to-end workflow for analyzing high-throughput phenotype microarray data in mammalian cell cultures. It begins with automated reading and noise correction, followed by comprehensive quality control via visual diagnostics, i.e. detecting missing wells, anomalous curve shapes and variance. Area under Curve is computed as a summary metric. Biological annotations (custom or in-built for PM-M01,PM-M02, PM-M11 - PM-M14) enable group-level comparisons.

MetPhenoR

The complete workflow including examples can be found in vignettes/MetPhenoR.html. A flowchart of the package can be found in man/figures/MetPhenoR_workflow_flowchart.png

Installation

MetPhenoR is available on CRAN:

install.packages("MetPhenoR")

Installation from GitHub

Install the latest version directly from GitHub remotes package:

# Install remotes if you haven't already
if (!require("remotes", quietly = TRUE)) {
  install.packages("remotes")
}

# Install MetPhenoR from GitHub
remotes::install_github(KatjaDanielzik/MetPhenoR)

Data Preparation

Before analyzing your data:

Create a new folder (e.g., Analysis).
Copy all your RawRead.csv and PlateInfo.csv files into this folder.
In PlateInfo.csv, create or select one column (e.g., Sample) to store experimental metadata.
Format entries like:
cell.line_variable1_variable_2
H460_treatment_2h
→ Use underscores, no spaces, and consistent capitalization (R is case-sensitive!).

Never use spaces — this causes errors in read_folder().

Key Features

Feature	Function
Noise correction: `OD(590nm) - OD(740nm)`	Removes background noise (valid for adherent cells only).
`qc_data()` plot	Spot missing files, inconsistent conditions, or technical issues early.
`classify_growth_curve_types()`	Identify flat, dropping, or noisy curves.
AUC + normalization	Compare across plates and conditions.
Built-in annotations	KEGG pathways (PM-M01/M02), drug info (PM-M11–M14).
Heatmaps & violin plots	Visualize biological trends.

Important Limitations

Not for bacteria or suspension cultures: OD(740nm) reflects cell growth in these systems.
AUC does not capture curve shape: two very different curves can have the same AUC.
Does not account for cell growth: if your cells grow rapidly, this may confound results.
Does not interpret biological meaning: a "hit" in a pathway does not imply activation.

Troubleshooting

Issue	Solution
`No data in qc_data()`	Check for spaces and variable names `Sample` column; ensure all files are in the folder.
`Filtering didn’t work`	Double-check `plate_type` and `filters` spelling.
`Plot is blank`	Check if `processed_data` contains data after filtering.