Interaction specification in SpinXML format currently also assumes a single common frame of reference for all spins and interactions. This is sufficient, but not necessarily convenient, particularly in solid state NMR, where chains nested Ibrutinib of reference frames (interaction → molecule → unit

cell → crystal → spinner → magnet) are often present, and in macromolecular NMR, where groups of spins belonging to different reference frames can move relative to one another. This calls for the creation of a reference_frame complex type shown in Fig. 7 and for the addition of a frame attribute citing the number of the relevant reference frame to rotation and vector complex types. Once the reference frame with id = “0” is defined as the laboratory frame, this amendment allows to specify chains and trees of reference frames, each with its own set of spins and interactions.

The resulting structure is illustrated in Fig. 8. It is particularly convenient in systems undergoing magic angle spinning or conformational dynamics; its elegance, however, is matched by the practical difficulty of implementing a parser, an export routine and an interactive editor for the resulting ultra-flexible format – we are therefore listing this feature as a possible extension that is not a selleck screening library part of SpinXML version 1.0 described in Section 3. Another minor limitation is the finite number of coupling specification styles in the interaction_term complex type (Fig. 1) – less common conventions (such as D, E specification for zero field splitting and “alphabet notation” for dipolar coupling) have not been included. Such design decisions are necessarily subjective and further specification styles could, if proven necessary, be added in future to either of the two SWITCH bars in the interaction_term complex type. Under any future expansion,

however, the SpinXML version 1.0 subset described in Section 3 will remain unchanged. The last noteworthy ifenprodil limitation of SpinXML is the absence of molecular dynamics (MD) variables, such as correlation times and order parameters. Although they could, particularly in protein NMR spectroscopy, be viewed as spin system parameters, they are not intrinsic to the spin system. MD parameters are also model-dependent and the number of models in the literature is unfortunately rather large. Critically, the models themselves and the community opinion on their relative merits continue to evolve, meaning that an attempt at standardization would be premature. The decision to not include correlation times and order parameters in SpinXML will be reviewed in due course – they may appear in the subsequent versions of the format that will be backwards compatible with the version described in Section 3.