Compare molecules with PharmCast

The two routes to a pharmacophore fingerprint, worked through on saquinavir and indinavir, two HIV-1 protease inhibitors of unrelated scaffold. The conventional route generates a conformer ensemble and runs the reference calculation over it. PharmCast predicts the same ensemble record from the two-dimensional structure and skips the ensemble entirely. The reference calculation puts the pair at a pharmacophore Tanimoto of 0.841 and PharmCast SCP v10 predicts 0.843, against a two-dimensional Morgan Tanimoto of 0.303. — The two routes to a fingerprint, and what this page runs, worked through on **saquinavir and indinavir** — two HIV-1 protease inhibitors of unrelated scaffold, and one of the examples below. The conventional route generates a conformer ensemble and runs the reference calculation over it. PharmCast predicts the same ensemble record straight from the structure. The reference calculation puts the pair at **0.841** and PharmCast SCP v10 predicts **0.843**, against a 2D Morgan Tanimoto of **0.303**. **saquinavir** `CC(C)(C)NC(=O)[C@@H]1C[C@@H]2CCCC[C@@H]2CN1C[C@@H](O)[C@H](Cc1ccccc1)NC(=O)[C@H](CC(N)=O)NC(=O)c1ccc2ccccc2n1` **indinavir** `CC(C)(C)NC(=O)[C@@H]1CN(Cc2cccnc2)CCN1C[C@@H](O)C[C@@H](Cc1ccccc1)C(=O)N[C@H]1c2ccccc2C[C@H]1O` Background: pharmacophore fingerprints, explained.

1 · The query molecule

Everything below is compared against this one.

You do not have to draw it. Paste a SMILES below and press Load — it is drawn for you.

Name it (optional)

2 · Compare it against

One molecule, drawn here, is enough. An address unlocks a list of up to 10 structures ranked against your query, and the report is then emailed to you as well as shown here.

You do not have to draw it. Paste a SMILES below and press Load — it is drawn for you.

Name it (optional)

Unlock a list · your email

The list, one SMILES per line, optional name after a space

up to 10 structures against the query

3 · Compare

Model

Try one of these. Every pair is two marketed drugs, found by scanning all 1,737 approved small molecules in ChEMBL for high pharmacophoric similarity with low 2D similarity. Numbers are what this page returns on SCP v10, the default model; selecting an earlier version changes them.

They act at the same target, through different chemistry

Amikacin vs paromomycin aminoglycoside antibiotics — 0.956 pharmacophorically, 0.456 by Morgan
Mometasone furoate vs alclometasone dipropionate corticosteroids — 0.690 against 0.361
Doxycycline vs demeclocycline tetracyclines — 0.820 against 0.481
Saquinavir vs indinavir HIV-1 protease inhibitors — 0.873 against 0.303
Formoterol vs olodaterol long-acting β₂-agonists — 0.877 against 0.300
Montelukast vs zafirlukast CysLT₁ antagonists — 0.472 against 0.114, the widest gap here

Different targets — a shared shape is not a shared activity

Prednisolone tebutate vs ixabepilone a corticosteroid and a microtubule inhibitor — 0.680 against 0.113
Dalfopristin vs cefditoren pivoxil a streptogramin and a cephalosporin — 0.545 against 0.104

Those last two are the lowest 2D similarities in the whole set, and they are listed apart on purpose: PharmSim says these molecules can present the same pharmacophores, which is a statement about shape, not about what they do.

PharmSim is the Tanimoto coefficient over the set pharmacophores of two ensemble fingerprints, so a set pharmacophore means a molecule can present that triangle of features in some accessible conformation, not that it always does. The 2D Morgan number sits beside each result for contrast, not as a calibration of it. Method, measurements and the training corpora: pharmacophore fingerprints, explained. To fingerprint a single molecule instead, use the fingerprint page.