Fit3D is a software for template-based and template-free detection of substructure similarity in biological structure data.
The biological function of proteins and nucleic acids, such as riboswitches or ribozymes, relies on the correct arrangement of small substructural units to catalyze substrates, bind ligands, or to preserve an ordered state. These molecular building blocks have evolved to retain similar interaction patterns and geometrical features to ensure functionality. While some of these patterns are reflected in recurring sequence motifs, evolutionarily remote proteins may only share a small set of similar structural motifs, which are not obvious at sequence level. The Fit3D algorithm is a highly accurate search engine to detect substructure similarities based on a given template. Additionally, it uses the mmm framework to discover structural motifs without the need for a template.
- search for template structural motifs with known function in protein structures of unknown function (protein function prediction)
- identify remote homologous structures by the identification of substructure similarity
- similar binding site screening for drug development or drug repurposing
- template-free detection of unknown structural motifs in a set of target structures
- Java 8 or later
For p-value calculation for Fit3D (optional):
- R installation 3.4.x or later
- local package installation privileges or the
Use the latest release provided as executable
An online version of Fit3D is available at biosciences.hs-mittweida.de/fit3d.
For detailed instructions on how to use the command line implementation please refer to the documentation.
If you use this software in your project, please cite:
Kaiser, F, Eisold, A, Bittrich, S, Labudde, D (2016)
Fit3D: a web application for highly accurate screening of spatial residue patterns in protein structure data.
Bioinformatics, 32, 5:792-4., doi:10.1093/bioinformatics/btv637