Welcome to the homepage of



ERNA-3D, winner of the Heinz-Billing prize 1997, is a Molecular Modeling System which was specially developed for the creation of models of large RNA molecules. The handling of proteins and other molecules is also possible.

With ERNA-3D it is very easy to build up a large number of double helices including single-stranded Internal Loops, Bulges and Hairpin Loops. For the creation, only a simple structure definition file is needed which contains information concerning the Primary Sequence and the Secondary Structure of the molecule (see fig. 1). The program translates the characters which define each residue of the primary sequence into spatial atomic coordinates (fig. 2). After that, it creates all the helical structures which are defined in the secondary structure (fig. 3 and 4).

The difference between ERNA-3D and conventional Molecular Modeling Systems is the ability to modify parts of molecules in a dynamic and realistic way. Single strands can be brought into a plausible configuration with the computer mouse. For this purpose a special Chain Translation Algorithm was developed which simulates all rotations of molecular groups around single bonds along the backbone chain of the molecule in real time. This gives the user the impression of watching a "realistic" molecule in front of him. Helices and parts of the molecule previously defined as a unit (Clusters) can be "clicked", translated and rotated, whereby all single strands connected to the structure will be pulled automatically in the same direction (see fig. 5).





In contrast to many other commercial programs, the relative arrangement of different molecules in one model can be changed very easily. All atomic models can be saved and loaded as PDB-files. The modified model and all user interactions can be examined three-dimensionally on the computer display with Shutter Glasses (necessary for 3-D viewing).

A major difficulty when modeling macromolecules is the preservation of clarity. If too many molecular parts are displayed simultaneously, it becomes nearly impossible for the user to work with the model. To avoid such difficulties, ERNA-3D allows the specification of selections, whereby only selected parts of the molecule are visible. Invisible parts remain stored in the memory of the computer, but cannot be modified by the user. Selections are also used to change the colors of molecular regions and to define Clusters. A cluster is the incorporation of different molecular groups into a rigid unit which can then be moved as whole.

ERNA-3D allows the generation of different degrees of molecular abstraction. As the crudest level of abstraction for a double helix, geometric cylinders can be created which then lie as "covers" over the helices (fig. 6a). An intermediate degree of abstraction is provided by tubes which are calculated along the backbone chain of the molecule (fig. 6b). The most detailed level of abstraction is given by CPK- and Ball-and-Stick-Models which can also be created by ERNA-3D (fig. 6c).



ERNA-3D has a logical architecture. The philosophy of ERNA-3D is that the modeling of molecules produces only two fundamentally different kinds of data. For this reason all data are strictly distinguished between atomic structures and their representations (polygon structures). Thus the most important lists in ERNA-3D are the atomlist and the polygonlist. This can be seen on the basis of the design of the graphical user-interface, where the most important pull-down menus are the Atomlist and the Polygonlist (fig. 7). Under Atomlist you find functions which can influence the atomlist in different ways. In an analogous manner, you find under Polygonlist all functions which influence the polygonlist. The user can switch very rapidly between the rendering of one such list or the rendering of both lists at the same time (mixed rendering of the atomic model and its representation).