Purpose of MOF+
The motivation behind MOF+ was mainly driven by the wish of ourselves to have a simple to use search engine to screen the vast amount of publicly available topologies on RCSR (and EPINET) in the particular way we need it for the RTA. Note that we not at all intend to become a "competitor" to the great RCSR page. All nets shown on MOF+ will have a clear indication with a direct link to were they come from. The vast majority of embeddings stems from RCSR, which is always clearly indicated. Recently we began to add derived embeddings by ourselves. If you are an experimentalist and crystallographer and you are mainly interested in the network deconstruction we actually recommend to have a look at RCSR, since there you might find the better and more complete descriptions. Our idea is to combine a rather elaborate and fine grained search on what type of vertex is connected to which, with our AAD classification scheme. In addition, we wanted to see the nets in 3D with the ability to directly rotate them on the screen.
On MOF+ MOF structures are not sorted by name (this can be ambiguous or hypothetical systems might have no name, yet) but by the name of the underlying net or embedding and the necessary building blocks. This implies that a MOF can appear twice, since there is no unique way to define the embedding and the building blocks. So if you see a MOF twice this is not a bug but a feature. Of course you can search by name not net name to filter. Most importantly the webweaver allows now to build new MOFs by selecting building blocks and embedding. This is currently restricted to the construction of one representative isoreticular isomer. Only for a selected set of BBs we have MOF-FF parameters available (the range of parameters is also constantly extended). In order to allow users to build initial structures of arbitrary systems we rely in the first step on an optimization with UFF4MOF (without charges assigned to the atoms).
We are working on a systematic extension of webweaver to allow for an automated GA screening of isoreticular isomers (see pages on RTA) and to make the MOF structure optimization more stable. For very large systems we observe various problems when assigning UFF4MOF parameters or during structural relaxation, which can make webweaver to fail. A further feature which is currently under development is the upload of experimental or externally computed structures (e.g. optimization of periodic MOFs with a periodic DFT code). In order to link such user uploaded structures into the database we need to detect the underlying net and the BBs.
Note that we are working on the page to extend it constantly with new features. Have a look at the Roadmap