assess_small_molecule
Read-only
Assess a small molecule given as SMILES or a molfile: RDKit's descriptors, medchem's drug-likeness rules (Lipinski, Veber, Ghose, Egan, REOS, rule of three, CNS and more) and the papers' own readings where medchem departs from them, with every bound and this molecule's value, a structural-alert screen (PAINS, Brenk, NIH, ZINC and ChEMBL's seven curated sets) with the matched atoms, the molecular formula with its masses and natural-abundance isotope pattern, QED drug-likeness under all three weightings, ESOL aqueous solubility, synthetic accessibility and natural-product likeness, the Bemis–Murcko scaffold and generic framework, the BOILED-Egg's gut-absorption and brain-penetration verdicts, and per-atom logP, polar-surface and charge contributions. Readings are taken with hydrogens implicit, the form the descriptors and the rules' papers assume. An alert flags a nuisance or reactive substructure; it is not a verdict on activity or toxicity.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| rules | No | which drug-likeness rules to evaluate, by id: medchem's 17, plus the papers' own readings of Lipinski, Ghose and REOS (ids ending _paper). Every rule by default. | |
| _files | No | The project: record references shaped like _subject. Their records become library documents a tool names in sequences/parts/document; a trace or alignment is found by file name. | |
| _subject | No | The record to act on: exactly one of sequence, accession (+source), sample (an id from list_samples) or file ({name, text} or {name, base64}); record_index picks one record of several. | |
| molecule | Yes | the structure: a SMILES string or an MDL molfile (V2000 or V3000). One molecule; for an SDF, pass one record. | |
| alert_sets | No | which structural-alert catalogs to screen against. Empty screens nothing, which is not a clean result. |
Output Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| qed | Yes | QED (Bickerton 2012) with mean weights, maximum weights and unweighted; the eight inputs (MW average mass, ALOGP Crippen, HBA, HBD, PSA, ROTB, AROM, ALERTS) and each input's desirability; and the SMARTS of every unwanted group QED counted | |
| rules | Yes | one entry per rule evaluated: id, label, passes, citation, and every clause with its property, bound and this molecule’s value. A rule with anyClauses passes when at least one of those holds | |
| alerts | Yes | every alert matched: its set, name, SMARTS and the atom indices of the first match, numbered on the implicit-hydrogen molecule | |
| formula | Yes | RDKit's molecular formula (every hydrogen, isotope labels merged, charge suffixed) with its element counts and charge, the monoisotopic and average masses in Da, and OpenMS's coarse natural-abundance isotope envelope (isotope index, mass, share of molecules), or the reason there is none (an isotope-labeled structure, or an element OpenMS has no isotope data for) | |
| summary | Yes | the assessment in one sentence, generated from this run | |
| scaffold | Yes | the Bemis–Murcko scaffold (ring systems and linkers, side chains removed) as canonical SMILES, empty for a molecule with no ring; the generic framework (every atom carbon, every bond single), null where an atom has more bonds than carbon allows; and the indices of the atoms the scaffold keeps, on the implicit-hydrogen molecule | |
| boiled_egg | Yes | the BOILED-Egg placement (Daina & Zoete 2016) of the neutral, unsalted ChEMBL parent as the paper places it (`placed` says whether that is the parent or, when the pipeline refuses one or it is the molecule itself, the record as drawn; `asDrawn` and `parent` carry both): tPSA counting S and P, RDKit's Crippen logP standing in for WLOGP, and for the gut-absorption (hia) and brain-penetration (bbb) ellipses whether the point is inside, its summed focal distance and the ellipse's major axis; region is yolk, white or outside | |
| solubility | Yes | ESOL (Delaney 2004) log S in mol/L under two coefficient sets, rdkitRefit (Walters' refit to RDKit descriptors) and delaney2004 (the paper's, with Crippen logP for Daylight CLOGP), and the four inputs: Crippen logP, average mass, rotatable bonds, aromatic proportion | |
| descriptors | Yes | RDKit's descriptors under RDKit's own keys (exactmw, CrippenClogP, tpsa, NumHBD …) | |
| np_likeness | Yes | Ertl et al. 2008 natural-product likeness, about -5 (synthetic-like) to 5 (natural-product-like), and the confidence: the share of the molecule's fragments the model knows | |
| canonical_smiles | Yes | RDKit's canonical SMILES of the molecule every reading was taken on | |
| atom_contributions | Yes | per-atom weights on the implicit-hydrogen molecule: logp (Crippen, heavy atoms), tpsa (Ertl, N and O), tpsaSP (Ertl with S and P) and charge (Gasteiger–Marsili; not a number around an element RDKit has no parameters for) | |
| bioavailability_radar | Yes | SwissADME's Bioavailability Radar (Daina 2017): six axes (lipo, size, polar, insolu, insatu, flex) with this molecule's value, the inclusive window, whether it is inside, and drugLike when all six are; read with Crippen logP for XLOGP3 and the XLOGP3 window's figure bound of 5.0 | |
| synthetic_accessibility | Yes | Ertl & Schuffenhauer 2009 synthetic accessibility, 1 (easy) to 10 (hard), with its fragment score, complexity penalty, symmetry correction, the counts they read and how many fragments the table did not know; null only if the fragment table could not be loaded |