Attachment Points and CAP Groups

A monomer is not a molecule but a fragment of one: it has open bonds where it joins its neighbours. This article explains how those bonds, the attachment points or R-groups, are written in a SMILES, what a cap group is, and why the service insists on both.

A monomer is a residue, not a free molecule

Alanine as a free amino acid is C[C@H](N)C(=O)O. Inside a peptide it has lost a hydrogen from its amine and a hydroxyl from its carboxylic acid, and it is bonded to two neighbours instead. The monomer registered in the library is that residue, with the two open bonds marked:

C[C@H](N[*:1])C([*:2])=O

Each [*:n] is a dummy atom standing for a bond to something else. The number is the R-group label: R1 and R2 here. HELM connects monomers by naming the R-groups on each side, so R2 of one residue bonds to R1 of the next.

The conventions behind the numbers

The numbers are a convention, not a chemical fact, and the tools that build sequences rely on it.

Polymer type	Monomer type	`R1`	`R2`	`R3`
`PEPTIDE`	`Backbone`	Amine nitrogen, N-terminal side	Carbonyl carbon, C-terminal side	Side chain, when the residue can branch, such as the epsilon amine of lysine or the thiol of cysteine
`PEPTIDE`	`Branch`	Bond to the backbone residue's `R3`
`RNA`	`Backbone` sugar	5' oxygen, to the previous phosphate	3' oxygen, to the next phosphate	1' carbon, to the base
`RNA`	`Backbone` phosphate	To the 3' of the previous sugar	To the 5' of the next sugar
`RNA`	`Branch` base	Bond to the sugar's `R3`
`CHEM`	`Undefined`	Free choice, but numbered from `R1` without gaps

Labels run from R1 to R25. They must start at R1 and increase without a gap: a monomer with R1 and R3 but no R2 is rejected.

Notations the service accepts

The service normalises every structure to the [*:n] form, and that is the form it shows when you reopen a monomer. It also reads:

Notation	Example	Where you meet it
Atom map on a dummy atom, `[*:n]`	`C[C@H](N[:1])C([:2])=O`	The recommended form, used by the import file and the form.
CXSMILES R-labels	`C[C@H](N[])C([])=O \\|$;;;_R1;;_R2$\\|`	Exported by chemistry toolkits; the label block after the `\\|` names each dummy atom.
Folded cap atom, `[H:1]`, `[OH:2]`	`C[C@H](N[H:1])C([OH:2])=O`	The original Pistoia HELM monomer library. The cap atom itself carries the map number.

[R1] as a bracket atom, and an unnumbered [*], are not attachment points. The first is not valid SMILES for the service; the second is a dummy atom with no label and is rejected with Number of R-Groups in Attachmentlist is NOT equal to number of R-Groups in Smiles.

Cap groups: what sits on an unused attachment point

An attachment point that is not bonded to a neighbour still has to be something. The cap group is the small group that completes the valence there: for a peptide backbone, a hydrogen on R1 gives the free amine, and a hydroxyl on R2 gives the free carboxylic acid. The cap is what the toolkit adds when it draws or computes a monomer at the end of a chain, or on its own.

The service stores the cap group as a name from a controlled vocabulary, and turns it into chemistry with a SMILES template where {n} is the R-group number:

Cap group	Template	Chemistry it completes
`H`	`[*:{n}][H]`	An amine, a thiol, an alcohol, an alkyl chain end. The default.
`OH`	`O[*:{n}]`	A carboxylic acid on `R2` of an amino acid, a phosphate oxygen.
`NH2`	`N[*:{n}]`	A C-terminal amide, a guanidine or amine attached through its carbon.
`Azide`	`[*:{n}]N=[N+]=[N-]`	An azido handle for click chemistry.
`Ethynyl`	`[*:{n}]C#C`	A terminal alkyne, the other partner of a click reaction.

Cap groups are written in the import file as R1:H;R2:OH and chosen from a selector per attachment point on the form. A cap outside the vocabulary is rejected with the list of valid values, never silently replaced: substituting a hydroxyl for an unknown cap would store a different molecule without anyone noticing. Your Discngine contact can add a cap group to your organisation's vocabulary when your chemistry needs one.

Why a wrong cap matters

The cap decides the molecule that tools compute from your monomer when the attachment point is free:

R2:H on an amino acid makes its free form an aldehyde, and the molecular weight and logP of every peptide ending on it are wrong;
R1:OH on an amino acid makes the N-terminus a hydroxylamine;
a cap chosen on the wrong side of a multi-atom template, such as writing the azide as if it were a single atom, gives a pentavalent carbon.

The cap does not change how the monomer bonds inside a chain. Two monomers with the same skeleton and different caps are the same residue in the middle of a sequence, and different molecules at its ends.

How to Create or Edit a Monomer: where the caps are set on the form.
Canonical SMILES and Duplicate Detection: why a change of cap counts as a change of the monomer.