From Protein Sequence to 3D Structure
When most people think about DNASTAR, they probably think of cloning and primer design in our Molecular Biology Suite or next-gen sequence assembly and analysis in our Genomics Suite. Indeed, much of our 30-year history as a company has been spent developing software in these domains. But some of our newest and most exciting developments are happening in the structural biology realm, specifically in protein structure prediction. With the ever-growing wealth of genomics data available, paired with the relatively low number of solved protein structures, there is an increasing need for software that can create 3D models of proteins without known structures. DNASTAR’s NovaFold, part of the Lasergene Structural Biology suite, allows scientists to do just that.
What is NovaFold?
NovaFold is based on I-TASSER, which has been ranked as the best algorithm for protein structure prediction in the Critical Assessment of Techniques for Protein Structure Prediction (CASP) competition for the past 10 years. The algorithm uses a unique hybrid approach that combines threading and ab initio modeling techniques to generate the best possible prediction model. While the homology modeling method used by most structure prediction software tools requires a high degree of sequence similarity between the query sequence and the template structure, NovaFold’s protein threading considers sequence similarity as well as predicted structural properties. Factors such as secondary structure, solvent accessibility and internal residue contacts all contribute to the selection of template structures, making it feasible to predict new structures, even without a high degree of sequence conservation. And unlike other modeling programs, NovaFold uses multiple template structures, making it possible to combine several domains into a single composite structure.
With the latest release of NovaFold, users can now define their own template structures. We’ve also added user restraint controls, allowing scientists to define properties such as known atomic distances or internal residue contacts. This flexibility allows researchers to refine the alignment and modeling for more tailored prediction results.
NovaFold is available in multiple configurations to meet the demands of today’s scientists and to efficiently utilize computing resources. NovaFold can run on the DNASTAR Cloud, hosted by Amazon, on a private service to protect user data. For those who prefer or need to keep all data in-house, we also offer a local version that runs on a Linux Workstation.
All of the functionality in NovaFold can be accessed from the easy-to-use Protean 3D application within the Lasergene Structural Biology Suite. Here, users can visualize predicted models, see predicted protein function and binding sites, and align predicted models to template structures.
Want to learn more about NovaFold? View our previous webinar here to see a demonstration of structure prediction in NovaFold and hear about some of the most important features in the software.