Ring closures

Ring closures let you bond non-adjacent atoms using SMILES-style digit suffixes. Two atoms with the same digit become bonded.

Basic syntax

A digit after an atom opens or closes a ring:

Source Rendered

C1-C-C-C-C-C1

$\mathrm{C}1-\mathrm{C}-\mathrm{C}-\mathrm{C}-\mathrm{C}-\mathrm{C}1$

This is cyclohexane: the first C1 and the last C1 are bonded, closing the ring.

Multiple rings

Different digits open different rings:

Source Rendered

C1-C2-C-C1-C2

$\mathrm{C}1-\mathrm{C}2-\mathrm{C}-\mathrm{C}1-\mathrm{C}2$

Rings 1 and 2 are independent.

Multiple closures on one atom

An atom can open/close multiple rings simultaneously:

Source Rendered

C12-C-C1-C2

$\mathrm{C}12-\mathrm{C}-\mathrm{C}1-\mathrm{C}2$

This opens rings 1 and 2 on the first atom, closes ring 1 on the third atom, and closes ring 2 on the fourth.

Rings inside groups

Ring closures work inside groups:

Source Rendered

(C1-C-C-C-C-C1)

$\left(\mathrm{C}1-\mathrm{C}-\mathrm{C}-\mathrm{C}-\mathrm{C}-\mathrm{C}1\right)$

((C1-C-C1))

$\left(\left(\mathrm{C}1-\mathrm{C}-\mathrm{C}1\right)\right)$

Ca(C1-C-C1)

$\mathrm{Ca}\left(\mathrm{C}1-\mathrm{C}-\mathrm{C}1\right)$

The group structure is preserved alongside the ring bond through CML round-trip.

CML output

Ring closure bonds appear as regular <bond> elements in CML, connecting the ring-closure atoms. The digit itself is preserved via the aci:ringClosures attribute so the round-trip is lossless:

<atom id="a1" elementType="C" aci:ringClosures="1"/>
...
<atom id="a6" elementType="C" aci:ringClosures="1"/>
<bond id="b6" atomRefs2="a1 a6" order="S"/>

Linter

The UnclosedRingCheck linter catches ring closure digits that have no matching partner:

Source Rendered

C1-C-C

$\mathrm{C}1-\mathrm{C}-\mathrm{C}$

This produces: [error] Atom(C): Atom has unmatched ring closure digit(s) "1".