Tools · Cheminformatics & Structure

Scaffold Analyzer

Group a hit list by Bemis–Murcko scaffold, count each framework, and pick one representative per scaffold.

Updated October 1, 2026

On this page

Prices, workflows, and method papers Open in the app

The Scaffold Analyzer shows which core frameworks a set of molecules is built on. It reduces each molecule to its Bemis–Murcko scaffold (ring systems and the linkers between them, side chains removed), counts how often each scaffold occurs, and picks a representative for each one. Use it to see whether a hit list is many variations on a few chemotypes or genuinely diverse. It runs on CPU with RDKit.

How it works

scaffold_type: murcko (the default; atom and bond types kept) or generic (all atoms carbon and all bonds single, which merges scaffolds that differ only in heteroatoms or saturation). Both forms are written for every molecule; this choice sets the grouping.
pick, the representative per scaffold: first (the default), highest_score (needs scores), or closest_to_scaffold (the molecule with the fewest extra atoms).
compute_mcs: the maximum common substructure between molecules, for small sets (up to 50 molecules).

Inputs

Up to 10,000 SMILES, with optional scores per molecule (as a scores list or a table) for picking the best representative. Convert SDF to SMILES first with Molecule Conversion.

Outputs

results.csv: each molecule's Murcko and generic scaffold, how many molecules share its scaffold (scaffold_count), and whether it is the representative (is_representative), plus pairwise MCS results when you asked for them.
The representatives as a SMILES set, for the next step.

For similarity-based clustering use Fingerprint Similarity & Clustering. For R-group analysis on a shared scaffold use Free Wilson Analysis.

Run it from the API

Submit with Submit a job and the job_type below. Price it first with Estimate job reservation cost: submitting reserves that amount from your wallet, and the charge settles at the actual runtime.