CNS Property Inspector and MPO Methodology Reference
Inspect a lorlatinib reference structure and the descriptors this workspace actually calculates. No CNS MPO score or blood–brain-barrier permeability prediction is produced here. The methodological discussion explains the six-input CNS MPO framework and identifies the logD and pKa information that cannot be inferred from the current descriptor panel.
Exact parent SMILES from the shipped ChEMBL 37 library. The preset provides molecular weight, Crippen logP, TPSA and hydrogen-bonding descriptors. It contains no fitted brain-exposure model, measured brain/plasma ratio or computed six-component MPO score.
Understand what an optimisation score combines
The CNS MPO approach combines several property desirabilities instead of applying one hard threshold to each molecule. Its purpose is to organise trade-offs across a specified set of physicochemical inputs. The six inputs are not redundant: neutral-species partitioning, distribution at physiological pH and basic pKa describe related but distinct aspects of ionisation and lipophilicity.
This workspace calculates some graph-derived descriptors but not all six required inputs. The missing values cannot be replaced by zero or assumed equal to an available quantity. In particular, substituting logP for logD ignores ionisation. A transparent property table with missing information is scientifically preferable to a complete-looking numerical score whose assumptions are concealed.
Frame a useful next experiment
Inspect the displayed molecular weight, polar surface estimate, donor counts and lipophilicity estimate as descriptive starting points. Then identify what additional measurements would address the actual project question. Passive permeability, transporter-mediated efflux and unbound brain exposure are different endpoints; a favourable result in one setting does not automatically answer all three.
When comparing an analogue series, preserve the method and conditions of every external value. Experimental pKa, calculated pKa and a value inferred from a related scaffold should remain distinguishable. The route can help organise those questions, but it does not fill the evidence gaps. Any future implementation of the six-input score would need an explicit method, input provenance and independent numerical reference cases before the result could be presented as calculated here.
Equations and variables
CNS MPO framework, described but not calculated
MPO = Σ di(pi), i = 1…6
- pi
- One of logP, logD7.4, molecular weight, TPSA, HBD or basic pKa
- di
- Property-specific desirability function defined by the cited method
- MPO
- Sum of six desirabilities; not an output of this workspace
Partition coefficient definition
log P = log10([solute]octanol/[solute]water)
- P
- Partition coefficient of the neutral species under the defined conditions
- [solute]
- Equilibrium concentration in the indicated phase (same concentration unit in both phases)
Scope and limitations
- No MPO result is calculated. The app lacks the required logD7.4 and basic-pKa models, and a partial sum would not be the cited six-input score.
- A descriptor profile or MPO score is not a quantitative prediction of BBB permeability, unbound brain exposure, transporter liability or clinical CNS efficacy.
- Changing the structure does not update missing experimental properties. Known behaviour of the reference compound must not be transferred to an edited analogue.
Sources and provenance
- RDKit descriptor implementation and versioned reference checksBSD-3-Clause
Method implementation; computed molecular properties are estimates.
- ChEMBL 37 parent structures in the shipped compound libraryCC BY-SA 3.0; share-alike
Existing library capture: 28 September 2026. Preserve attribution when redistributing its rows.
- Wildman and Crippen: atomic contributions to molecular propertiesCited method; no reproduced article data
- Ertl, Rohde and Selzer: fast calculation of molecular polar surface areaCited method; no reproduced article data
- Wager et al.: moving beyond rules with CNS multiparameter optimisationCited method; no article dataset redistributed
Primary paper consulted on 16 October 2026. Six inputs include logD at pH 7.4 and basic pKa.
Technical questions
Where is the CNS MPO score?
It is not calculated by the current engine. This route opens an existing descriptor inspector and states that limit before the methodology. Supplying four available properties and silently replacing the missing logD and pKa values would produce a different, unsupported score.
Does a favourable profile show that my compound crosses the BBB?
No. Brain exposure depends on more than these descriptors, including ionisation, transport, binding and systemic exposure. A property framework can guide optimisation, but a permeability or brain-exposure claim requires an appropriately qualified model or experimental evidence.
Why use lorlatinib as the reference?
It is a named, cited structure already present in the local compound library, so the graph can be inspected immediately. The reference does not grant the workspace a brain-exposure model, and editing the graph does not establish that an analogue preserves the original compound’s pharmacology.
Connected workflows
- TPSA and Crippen logP calculation
Review the available polarity and lipophilicity descriptors.
- Lipinski rule-of-five property inspection
Contrast a property heuristic with an exposure claim.
- EGFR sequence and structure inspection
Inspect target evidence without confusing it with brain exposure.