MCDL

The Modular Chemical Descriptor Language (MCDL) is a compact linear notation for representing molecular structures and associated compound data. It was introduced in 2001 by Andrei A. Gakh and Michael N. Burnett (Oak Ridge National Laboratory) in the Journal of Chemical Information and Computer Sciences, with stereochemistry modules added in a 2011 follow-up in the Journal of Cheminformatics. Like other line notations it belongs to the same family as SMILES, InChI and InChIKey, but its distinguishing feature is an explicitly modular design: a molecule is described by a sequence of self-contained modules rather than a single monolithic string.

Two modules are mandatory and carry the molecular graph. The composition module lists atoms and small atom-groups (fragments) in a semicolon-separated, canonically ordered sequence — for ethanol, roughly "C;CHH;2CHHH;CO;2OH" — placing hydrogen counts directly on their heavy atoms rather than in a separate layer as InChI does. The connectivity module then encodes bond topology as a compact per-atom adjacency list of neighbor indices and bond orders, e.g. "[2,3,5,6;4;;;7]". Because both mandatory modules are written in canonical form, two MCDL strings can be compared directly for identity and substructure matching without re-canonicalization. Separate optional modules cover tetrahedral atom and double-bond stereochemistry, and further supplementary modules can carry names, 2D/3D coordinates, elemental analysis, mass/NMR spectra and physical-chemical properties, so the notation doubles as a lightweight record format.

In the graph-data and conversion world MCDL is a coordinate-free molecular graph encoding, contrasting with connection-table formats such as Molfile, SDF and MOL2 that store explicit Cartesian coordinates. Its reference implementation is a GPL Java structure editor (~200 KB applet) that regenerates publication-quality 2D diagrams from connectivity alone, using a template database of common ring systems, iterative fragment matching, overlap resolution and Kekule reconstruction for aromatic rings. MCDL round-trips to and from these formats through Open Babel (Open Babel), which has bundled MCDL read/write support since version 2.3.1.

Its strengths are a genuinely canonical, human-readable string that keeps hydrogen and fragment information local, plus an extensible module system for attaching properties. Limitations are practical rather than technical: adoption stayed narrow, it never gained the ecosystem momentum of SMILES or the registry role of InChI, tooling beyond the original editor and Open Babel is sparse, and the notation is best suited to organic small molecules rather than polymers or macromolecular structures.