Ketamine
Targets in the model
4
each with a measured Kᵢ
A one-compartment model of what Ketamine might do over time, at a dose and by a route you choose. The affinities driving the occupancy calculation are measured and this archive holds them; the pharmacokinetics are mostly estimated, and the table at the foot of the page says which is which. The measured dossier · method
The input
First-order gastrointestinal absorption with a rate constant giving a peak near one hour, and a bioavailability reduced by first-pass metabolism. Both are class placeholders; where a compound has a published oral F it overrides this. Bioavailability for this route is 0.3 and the absorption rate constant 1.2 h⁻¹, both properties of the route rather than of Ketamine.
Modelled concentration
3 h 36 min
One compartment with first-order absorption. It cannot show a distribution phase, so the first minutes after an intravenous dose are the least trustworthy part of the curve. Peak effect-site concentration is 42 nM.
Reported milestones
No entry. Controlled subanesthetic studies in healthy volunteers describe the dissociative and psychotomimetic effects in detail — Krystal et al. 1998 (PMID 9498724) rates them on the BPRS and CADSS during a bolus-plus-infusion regimen — but report them by rating scale rather than against a clock, and a milestone with no time cannot be placed on a time axis. Ketamine's sandbox therefore shows model landmarks only.
Model landmarks
These are not observations. Each one is a feature of the curve above, read off the figure and moving with it whenever you change a parameter.
Dose given0
The model starts here.
Peak plasma1 h 35 min
Modelled maximum of 77.4 nM total plasma concentration.
Peak effect site1 h 46 min
Modelled maximum of 42.3 nM at the effect site — the concentration the occupancy below is computed from.
Effect site down to a tenth11 h 2 min
A threshold this model picked, not a clinical endpoint. A first-order curve never reaches zero.
Modelled receptor occupancy at 3 h 36 min
O = C/(C+Kᵢ) at the effect-site concentration above. The Kᵢ is measured and this archive holds it; the C is modelled. Equilibrium is assumed at every instant, with no receptor reserve, no active metabolite and no tolerance.
This archive holds no pChEMBL value for the pair, and holds one for others of this compound — so the absence is ChEMBL’s, not a gap in the harvest. ChEMBL withholds a pChEMBL where the activity is not a comparable, meaningful result: an inexact relation, an unrecognised type, or a screen the depositor marked inactive. The concentration is shown because it is what the archive holds, and it is ranked below everything vouched for because it is not evidence of binding. Ketamine against the APE1 endonuclease is the example that earned this footnote: ChEMBL records that screen as inactive, and read as an affinity it made a DNA-repair enzyme ketamine’s tightest binder by two orders of magnitude.
Modelled regional engagement
Occupancy at each target multiplied by that target’s transcript level in the region, summed. This product is the model speaking. A dossier’s Systems plate deliberately refuses to compute it, because there it would read as a finding; here it is an output with its assumptions on the page. Transcript is not receptor density, and a region shaded here is a region where the model says binding could occur, not one shown to be affected.
And nothing checks it. The concentration plot above carries reported milestones because human studies state times; this figure has no counterpart. The human imaging work this archive cites reports connectivity change across a cortical parcellation, which is a different quantity in a different set of regions — the atlas holds the cortex as one — so it can neither confirm nor contradict the shading here, and translating it onto these thirteen regions would assert a correspondence nobody has published.
Brain
Body
Shading is on a fixed scale — the strongest engagement reached anywhere during the whole run — so the figures dim as the compound clears instead of renormalising to whatever the current instant happens to hold. Outlines from the EMBL-EBI Anatomogram, the same plates the dossiers use.
Brain regions
- cerebellum
- basal ganglia
- hypothalamus
- cerebral cortex
- amygdala
- retina
- hippocampal formation
- midbrain
- spinal cord
- pituitary gland
Peripheral & autonomic
- skeletal muscle
- adrenal gland
- thyroid gland
- spleen
- salivary gland
- stomach 1
- lung
- colon
- pancreas
- duodenum
- small intestine
- smooth muscle
Parameters
The whole model is these five numbers plus the route. Each says whether anyone measured it for Ketamine. Edit any of them; nothing here is written back to the archive.
How long the plasma concentration takes to halve once absorption is over.
Class prior narrowed to the commonly reported 2-3 h terminal half-life; this archive holds four T1/2 rows for ketamine but none records species or assay conditions.
Apparent volume the dose would occupy at the plasma concentration. Large means the compound leaves the blood for tissue.
Class prior, slightly reduced; ketamine is lipophilic but less so than the cannabinoids.
The share of plasma drug not bound to protein. Only unbound drug crosses into the brain.
Ketamine is reported to be modestly protein-bound. The archive holds three PPB rows without conditions, so this is not taken as measured.
Unbound brain concentration divided by unbound plasma. Below 1 implies active efflux; 1 is passive equilibration.
Passive equilibration assumed.
How fast the receptor compartment follows plasma. Small values put the peak effect long after the peak concentration.
Raised above the prior: intravenous ketamine has a notably fast onset, so the effect site equilibrates quickly.