Multi-Modal Foundation · openadmet
PXR Activation Model
A baseline CheMeleon model predicting pregnane X receptor activation potency to assess drug-drug interaction risks.
What PXR Activation Model is
The PXR Activation Model (openadmet/pxr-chemeleon-baseline) is a baseline computational tool developed by openadmet to predict the activation of the pregnane X receptor (PXR).
By predicting half-maximal effective concentration values in logarithmic form (pEC50) for candidate chemical compounds, the model aims to support early-stage identification of drug-drug interaction liability in drug discovery workflows. However, the developers explicitly caution that model performance is poor, making it intended primarily as a baseline reference.
How it works
The model is a single-task CheMeleon architecture trained on public pEC50 data curated from ChEMBL for the PXR target. It predicts mean pEC50 activity values from input molecular structures provided as canonical SMILES strings.
Key facts
- Architecture
- Single-task CheMeleon
- Target
- Pregnane X receptor (PXR)
- Prediction target
- pEC50
- DOI
- 10.57967/hf/7795
What you can use it for
- Early drug-drug interaction liability screening
- Screening molecular chemical libraries to identify potential PXR activation and nuclear receptor-mediated CYP induction.
- Baseline comparison for ADMET modelling
- Serving as an open baseline reference model when developing novel machine learning architectures for PXR activity prediction.
- Batch compound scoring
- Predicting pEC50 values from canonical SMILES strings for collections of compounds using the openadmet command-line framework.
Run it here
We list this model but cannot run it yet: no pipeline_tag — needs a human.
Everything on this page still applies. Running it needs a loader for this repository, not a bigger machine.
Strengths and limits
Good at
- Targets a critical mechanism of drug-drug interaction liability via nuclear receptor-mediated CYP enzyme induction.
- Provides a reproducible, open baseline using public ChEMBL biological data.
- Easily integrates into compound inference pipelines via the openadmet command-line interface with support for CPU and GPU acceleration.
Watch out for
- Model performance is documented as poor, and outputs should be treated with caution.
- Outputs empty standard deviation columns as it is a single model rather than an ensemble, meaning it cannot provide uncertainty estimation.
Inputs and outputs
You provide
Sequence or text
A protein or nucleotide sequence, or free text — whichever this model was trained on.
You get back
Mean-pooled representation, one vector per input.
Licence
Check the licence before commercial use · apache-2.0
The apache-2.0 licence carries terms worth reading before you depend on it.
Questions
Can I use PXR Activation Model commercially?
Check the licence before commercial use. The apache-2.0 licence carries terms worth reading before you depend on it.
What does the model predict?
It predicts the mean pEC50 activation potency value (chemprop-chembl_pchembl_value_mean) of chemical compounds against the pregnane X receptor.
Why are the standard deviation columns empty in the output?
The model is a single-task baseline model; uncertainty estimation requires training an ensemble of models, so standard deviation values cannot be calculated.
What input format does the model require for predictions?
Inference requires a CSV file containing canonical SMILES representations of molecules (such as under the OPENADMET_CANONICAL_SMILES column).
How to cite it
10.57967/hf/7795
Written from PXR Activation Model's own documentation (model card). Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.