DSSR deliberately makes a distinction between ‘stem’ and ‘helix’, as shown below:

a helix is defined by base-stacking interactions, regardless of bp type and backbone connectivity, and may contain more than one stem.

a stem is defined as a helix consisting of only canonical WC/wobble pairs, with a continuous backbone.

By definition, a helix or stem consists of at least two base-pairs with stacking interactions. Helix is more inclusive and may contain more than one stem. This differentiation between ‘helix’ and ‘stem’ naturally leads to the definition of coaxial stacking, another widely used yet vaguely specified concept.

Again, the abstract notion can be best illustrated with a concrete example. In the classic yeast phenylalanine tRNA (PDB id: 1ehz), DSSR identifies that two stems [the acceptor stem (right) and the T stem (left)] are coaxially stacked within one double helix. See the figure below.

In the above schematics cartoon-block representation, each Watson-Crick base pair is rendered as a single, long rectangular block. Base identities of the G–U wobble, and the two non-canonical pairs (left terminal) are illustrated separately, with a larger block size for purines (G and A), and a smaller size for pyrimidines (C, U, and T).

I picked up ‘stem’ as a more specialized duplex because it is widely used in the RNA stem-loop structure, and in describing the four ‘paired regions’ of the classic tRNA cloverleaf secondary structure. On the other hand, ‘helix’ is (to me at least) a more general term, and thus more inclusive. It is worth noting that other terms such as ‘arm’, ‘paired region’, or ‘helix’ etc. have also been used interchangeably in the literature to refer what DSSR designated as ‘stem’.

As a side note, the basic algorithm for identifying helixes/stems in DSSR is also applicable for detecting G-quadruplexes. The same idea of ‘helix’ or ‘stem’ also applies here (see figure below for PDB entry: 5dww). Indeed, as of v1.7.0-2017oct19, DSSR contains a new section for the identification and characterization of G-quadruplexes.

DSSR is “an integrated software tool for dissecting the spatial structure of RNA”. It excels in consolidating the diverse pieces together via a coherent framework, readily accessible in a solid software product. DSSR may well serve as a cornerstone in RNA structural bioinformatics and would facilitate communications in the broad areas related to nucleic acids structures.