CYP Data Engine

Code and data to reproduce the figures in the CYP Data Engine blog post.

Prerequisites

• Quarto (>= 1.4)
• R (>= 4.5) with renv
• Python (>= 3.13) with uv

Setup

R packages — restore the renv lockfile:

renv::restore()

Python packages — create the virtual environment from the lockfile:

uv sync

Rendering

Blog version (clean, no code shown):

quarto render

Output: _site/index.html

Notebook version (code visible with fold/expand):

quarto render --profile notebook

Output: _site/notebook/index.html

Protocols

Experimental protocols used to generate the datasets in this blog post:

• CYP Reactivity Assay — CYP3A4/CYP2J2 reactivity screening via SCIEX Echo MS+ ZenoTOF 7600
• CYP3A4 Inhibition Assay — CYP3A4 inhibition (active preincubation) via fluorescence readout

Data

All data is in data/:

File	Description	Rows
reactivity.tsv	Raw Echo-MS peak areas for CYP3A4 and CYP2J2 reactivity screening (control vs treatment, per compound/replicate)	19,344
inhibition.tsv	CYP3A4 pIC50 values, QC flags, and SMILES for inhibition screening	1,340
inhibition_wells.tsv	Well-level fluorescence data for CYP3A4 inhibition DRC fitting	16,931
willitfly.tsv	Ionization comparison: 1 mM ammonium fluoride vs 5 mM ammonium formate buffers	11,353
clearance_snippet.tsv	Example metabolic clearance time-course data	292
tdi_drc_curves.tsv	Fitted dose-response curves for time-dependent inhibition (TDI) assay	3,012
tdi_drc_params.tsv	Fitted DRC parameters (pIC50, Hill slope, CIs) for TDI compounds	12
tdi_drc_points.tsv	Raw DRC measurement points for TDI compounds	144
tdi_pic50_shift.tsv	pIC50 shift estimates (±preincubation) for TDI compounds	6
inhibition_drc-plots/	Dose-response curve images (1,343 PNGs) used as tooltips in Figure 3	—