# Chemical probe

Source: https://onco.cc/terms/chemical-probe/  
OnCo record `chemical-probe` (Term). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

A chemical probe is a compound used in the laboratory to switch one biological process off or on and ask what it does, so it is judged by how well it is characterised for that question and not by whether it could be given to a person.

## Summary

A chemical probe is a research compound used to perturb a defined biological process in cells or animals. It is useful only to the extent that it is potent enough, selective enough and characterised well enough for the question being asked. Arrowsmith and colleagues put the requirement as a probe that is potent, of known selectivity and with a proven mechanism of action, and describe how many probes in common use were first characterised inadequately and later proved to be nonselective. LY294002 was described in 1994 as a selective inhibitor of PI3 kinase and was still sold as one by nearly all vendors, although by 2005 it was clear that it inhibited many other proteins at the concentrations used to inhibit PI3 kinase. An effect seen in cells at a high concentration therefore identifies an effect at that concentration, not necessarily a target.

The numerical standards belong to the groups that set them. For probes of epigenetic targets, the Structural Genomics Consortium as described by Arrowsmith and colleagues asks minimally for in vitro potency at the target below 100 nM, more than 30-fold selectivity over related proteins of the same family, profiling against a standard panel of off-targets and effects on the target in cells below 1 uM, and now requires an inactive close analogue as a negative control. Workman and Collins advocate assessing fitness for the intended experiment rather than applying an overly restrictive universal rule. Those cut-offs were written for particular target classes and do not apply to every probe, and neither paper offers a number that makes a concentration universally clean.

Controls carry the causal argument. An inactive or weakly active analogue from the same chemical family asks whether the effect follows the target-binding feature rather than shared chemistry; an active probe of a different chemical structure asks whether an independent perturbation agrees, with the limit that two probes can still share an off-target; and a genetic change or engineered resistance asks whether altering the target alters the effect as predicted, with the limit that removing a whole protein is different from blocking one of its functions. Workman and Collins add that a chemical probe can inhibit a specific function rather than remove the protein, which RNA interference cannot, and that both approaches have off-target as well as on-target effects. Agreement among these tests strengthens an inference, and conflicts should be recorded rather than set aside.

A probe is not a medicine. Arrowsmith and colleagues note that a probe does not need to meet the requirements of a successful medicine, such as good pharmacodynamics and oral bioavailability, and that high-quality probes have served as seeds for new medicines. A medicine also needs evidence in people about product quality, exposure, tolerability and benefit for its intended use, so a useful probe may lack adequate delivery, stability or safety, and an approved medicine can be a poor tool for isolating one target if several of its interactions contribute to its effects; this is an evidence-reading distinction rather than a claim that the same experiment establishes clinical suitability. A culture concentration is not a patient dose.

Target engagement is the separate question of whether a compound interacts with its target where the experiment needs it. The cellular thermal shift assay, introduced by Martinez Molina and colleagues, is built on ligand-induced thermal stabilisation of target proteins and was used to monitor drug binding in cells and tissue samples; a shift can support engagement under the assay conditions and is not by itself an inhibition or occupancy percentage. Engagement, a change in target function and cell survival are three measurements, and an effect on viability does not identify the responsible target.

Worked example, fictional. A new compound lowers the viability of cultured cells at 10 micromoles per litre, and neither engagement nor selectivity has been measured. The supported statement is that the compound has an effect in this culture at this concentration. It does not show that a particular protein is essential, and it does not show that a safe human dose reaches that protein. To close those gaps one would measure engagement and relevant selectivity, test an inactive analogue and a chemically different probe, and perturb the target genetically or by engineered resistance. Those steps address causality; exposure, quality, safety and clinical evidence remain separate questions.

## Fields

- Kind: Term
- Last checked: 2026-11-11
- Also known as: chemical probes; research probe compound; probe compound; chemical probe versus drug

## Notes

- Record the chemical identity, structure, source and purity, concentration-dependent activity, selectivity and effects in the relevant system. An inactive close analogue is useful, but its uptake or other properties may differ; an unrelated active probe can still share an off-target. Removing a whole protein can differ from inhibiting one function. Agreement among controls strengthens an inference in that system, and disagreement belongs in the evidence record rather than being discarded.

## Sources

- Arrowsmith et al., the promise and peril of chemical probes (2015): https://pubmed.ncbi.nlm.nih.gov/26196764/
- Workman and Collins, probing the probes: fitness factors for small molecule tools (2010): https://pubmed.ncbi.nlm.nih.gov/20609406/
- Martinez Molina et al., monitoring drug target engagement in cells and tissues using the cellular thermal shift assay (2013): https://pubmed.ncbi.nlm.nih.gov/23828940/

## Connected records

- terms: [Cyclin-dependent kinase inhibitors](https://onco.cc/terms/cdk-inhibitors/), [DNA damage response inhibitors](https://onco.cc/terms/dna-damage-response-inhibitors/), [Drug target and mechanism of action](https://onco.cc/terms/drug-target-and-mechanism/), [Small molecule drug](https://onco.cc/terms/small-molecule/), [Therapeutic index (therapeutic window)](https://onco.cc/terms/therapeutic-index/)
- targets: [BRD4](https://onco.cc/targets/brd4/)

---
JSON: https://onco.cc/api/v1/entities/chemical-probe.json