Drug Discovery · openadmet
openadmet-expansionrx-challenge-data
ADMET profiling data for RNA-targeted small molecules across multiple pharmacokinetic endpoints.
What openadmet-expansionrx-challenge-data is
This dataset compiles empirical absorption, distribution, metabolism, excretion, and toxicity (ADMET) measurements from preclinical drug discovery programmes conducted by Expansion Therapeutics. The underlying chemistry focuses on small-molecule compounds developed to target RNA-mediated diseases.
Originally assembled for the OpenADMET-ExpansionRx blind challenge, the collection provides a benchmark for evaluating computational methods in predictive pharmacokinetics. It reflects realistic medicinal chemistry scenarios where models trained on earlier discovery iterations are tasked with anticipating the property profiles of late-stage candidate molecules.
How it was made
The data reflects compound profiles collected during lead optimisation programmes. Compounds are provided in two distinct configurations: an ML-ready dataset that retains only in-range numerical assays, and a raw dataset that includes censored assay outcomes containing out-of-range measurement indicators, such as inequality signs.
The core evaluation spans nine experimental endpoints: lipophilicity (LogD), kinetic aqueous solubility (KSOL), intrinsic clearance in mouse and human liver microsomes (MLM and HLM CLint), permeability and efflux ratio in Caco-2 cell monolayers, and the unbound fraction in mouse plasma, brain, and gastrocnemius muscle tissues. An additional endpoint covering rat liver microsomal (RLM) clearance is supplied within the raw collection.
Key facts
- Data provider
- Expansion Therapeutics / openadmet
- Modality
- Tabular chemical and ADMET data
- Format
- CSV
- Core endpoints
- 9 (plus 1 supplementary endpoint in raw data)
- Target biology
- RNA-mediated diseases
- Digital Object Identifier
- 10.57967/hf/9687
What you can use it for
- ADMET Property Prediction
- Training regression and multi-task machine learning algorithms to predict physical and metabolic properties of small-molecule drug candidates.
- Lead Optimisation Benchmarking
- Assessing how effectively computational algorithms extrapolate from early campaign compounds to structurally refined, late-stage analogues.
- Censored Data Modelling
- Using the raw dataset to design and evaluate survival analysis or bounded regression models that account for out-of-range assay values.
Explore the data
The schema, real rows, full-text search and a SQL filter — read in your browser, nothing downloaded.
Explore
Read the real rows without downloading anything
Reading rows…
LiveRead from openadmet/openadmet-expansionrx-challenge-data at the moment you asked. Nothing is cached or stored — every row above came from that request.
How the columns are distributed
Over 5,326 rows of default/train
Caco-2 Permeability Efflux
float- min
- 0.26
- median
- 1.38
- mean
- 4.22
- max
- 105.64
59.4% empty
Caco-2 Permeability Papp A>B
float- min
- 0
- median
- 9.82
- mean
- 12.40
- max
- 51.41
59.5% empty
HLM CLint
float- min
- 0
- median
- 16.40
- mean
- 52.78
- max
- 2589.90
29.4% empty
KSOL
float- min
- 2.9e-3
- median
- 131
- mean
- 146.34
- max
- 325
3.7% empty
LogD
float- min
- -2
- median
- 2.10
- mean
- 2.11
- max
- 5.20
5.4% empty
MBPB
float- min
- 0
- median
- 3.38
- mean
- 7.67
- max
- 99.23
81.7% empty
MGMB
float- min
- 0
- median
- 4.42
- mean
- 7.53
- max
- 61.80
95.8% empty
MLM CLint
float- min
- 0
- median
- 219.65
- mean
- 560.53
- max
- 10354.80
15.1% empty
MPPB
float- min
- 0
- median
- 8.90
- mean
- 14.75
- max
- 87.60
75.6% empty
Molecule Name
string_text- min
- 9
- median
- 9
- mean
- 9
- max
- 9
SMILES
string_text- min
- 22
- median
- 46
- mean
- 48.03
- max
- 88
Strengths and limits
Good at
- Derived from authentic industrial lead optimisation campaigns rather than aggregated public literature.
- Supplies both an in-range tabular version for standard machine learning workflows and an uncurated raw version containing inequality qualifiers.
- Measures specialised distribution parameters, such as brain and muscle tissue protein binding, alongside conventional microsomal stability and solubility assays.
Watch out for
- The ML-ready version discards assay values that fall outside analytical measurement boundaries.
- The chemistry is centred on small molecules directed at RNA targets, which may restrict performance on disparate structural drug classes.
- Handling the raw records requires preprocessing to accommodate inequalities and assay limits.
Questions
Can I look inside openadmet-expansionrx-challenge-data without downloading it?
Yes. This page reads its schema, real rows and per-column statistics in the browser, with full-text search and a SQL filter. Nothing is downloaded to your machine.
What licence is openadmet-expansionrx-challenge-data under?
Its authors declare cc-by-4.0. Read the terms on the original page before reusing the data.
What is the distinction between the ML-ready and raw versions?
The ML-ready dataset contains solely in-range numerical values split into training and test sets. The raw version retains original out-of-range flags, such as greater-than signs, and includes an additional rat liver microsomal clearance endpoint.
Which biological clearance assays are recorded?
The benchmark tracks intrinsic clearance in human and mouse liver microsomes across both formats, with rat liver microsomal clearance recorded in the raw version.
Which tissue binding measurements are present?
The data records the percentage of unbound compound across three mouse matrices: plasma protein binding, brain protein binding, and gastrocnemius muscle binding.
Written from openadmet-expansionrx-challenge-data's own dataset card. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.