SMILES Notation

Shape

Length 1 to 1000 characters.

Told apart

Background

Origin

SMILES, the Simplified Molecular Input Line Entry System, is a line notation that uses printable characters for entering and representing molecules and reactions.^[1] Daylight's SMILES theory manual records that development of SMILES was initiated by David Weininger at the Environmental Research Laboratory of the U.S. EPA in Duluth, Minnesota.^[1] The SMILES design was completed at Pomona College in Claremont, California.^[1] The Environmental Protection Agency funded the initial project to develop SMILES.^[2] David Weininger's paper "SMILES, a Chemical Language and Information System. 1. Introduction to Methodology and Encoding Rules" was published in the Journal of Chemical Information and Computer Sciences in 1988.^[3] A second paper, "SMILES 2. Algorithm for Generation of Unique SMILES Notation", by David Weininger, Arthur Weininger and Joseph L Weininger, appeared in the same journal in 1989.^[3]

The OpenSMILES specification says that since its introduction in the late 1980s, SMILES has become widely accepted as a de facto standard for exchange of molecular structures.^[3] In 2007, an open standard called OpenSMILES was developed by the Blue Obelisk open-source chemistry community.^[2] The OpenSMILES authors wrote that Daylight's specification, being proprietary, limits the universal adoption of SMILES and has no mechanism for contributions from the chemistry community.^[3] The current OpenSMILES specification is version 1.0, dated 2016-05-15.^[3]

Use

Hydrogen atoms do not normally need to be specified when writing SMILES for most organic structures.^[1] In SMILES, atoms in aromatic rings are specified by lower case letters, so aromatic carbon is written c.^[1] SMILES branches are specified by enclosing them in parentheses, and can be nested or stacked.^[1] A SMILES ring closure is marked by a digit immediately following the atomic symbol at each end of the ring-closing bond.^[1] Disconnected compounds are written in SMILES as individual structures separated by a period.^[1] SMILES written with isotopic and chiral specifications are collectively known as isomeric SMILES.^[1]

There are usually a large number of valid SMILES strings which represent a given structure.^[1] A canonicalisation algorithm exists to generate one special SMILES among all the valid possibilities.^[1] RDKit's documentation shows the inputs C1=CC=CN=C1, c1cccnc1 and n1ccccc1 all being written back as the same canonical SMILES, c1ccncc1.^[4] A canonical SMILES depends on the canonicalisation algorithm used to generate it.^[2] The Daylight and OpenEye algorithms for generating canonical SMILES differ in their treatment of aromaticity.^[2] OpenSMILES warns that a canonical SMILES should not be considered a universal, global identifier, because two systems may use different rules and one system's output can change as it is improved or has bugs fixed.^[3] OpenSMILES advises that anyone considering canonical SMILES for a database system should also investigate InChI.^[3] OpenSMILES notes that names for SMILES flavours such as isomeric SMILES conflict between vendors and toolkits.^[3] When you join or deduplicate molecules from more than one source, re-canonicalise every string with one toolkit and one version, rather than comparing the strings as they arrived. (Meridian's view)

Under OpenSMILES, a SMILES string ends at a whitespace character (space, tab, newline or carriage return) or at the end of the string.^[3] OpenSMILES allows other data, such as a name, properties or a registration number, to follow the SMILES on a line after the whitespace.^[3] RDKit's MolFromSmiles returns None when it fails to parse a SMILES string, with an error message such as "Can't kekulize mol".^[4] Short values such as O, CC or CCO read as ordinary text, so trust a SMILES column on its header and on whether its values parse in a chemistry toolkit, not on a character pattern alone. (Meridian's view) Keep SMILES as case-sensitive text through every step of a pipeline, and if a line from a .smi file holds more than one field, split it at the first whitespace before parsing. (Meridian's view)

Where you'll meet it

ChEMBL's chembl_37 release publishes chembl_37_chemreps.txt.gz, a tab-separated file containing SMILES, InChI and InChI Key representations of its compounds.^[5] ChEMBL's COMPOUND_STRUCTURES table holds a CANONICAL_SMILES column, generated using RDKit.^[6] ChEMBL's schema documentation declares CANONICAL_SMILES as VARCHAR2(4000).^[6]

Sources

1. [1]Daylight Theory: SMILESdocumentation
2. [2]Simplified Molecular Input Line Entry System - Wikipediasecondary source
3. [3]OpenSMILES specificationstandard
4. [4]Getting Started with the RDKit in Python — The RDKit 2026.09.1 documentationdocumentation
5. [5]ChEMBL FTP Directorydataset
6. [6]chembl_37 Schema Documentationdataset

Detection notes

Use it

Schema

Aliases