Protein 3D structure prediction from a unique amino acid sequence is so important because having a protein structure model provides a greater level of understanding of how a protein works, which can allow us to create hypotheses about how to affect it, control it, or modify it. Many tools for protein structure prediction rely solely on homology modeling, which only works well for proteins that have a high degree of sequence similarity to protein sequences in the Protein Data Bank.
By contrast, NovaFold, NovaFold AI & NovaFold AI-Multimer use three different award-winning algorithms to create highly accurate, full 3D atomic models of proteins that are unattainable through standard modeling methodologies.
NovaFold AI uses the AlphaFold 2 algorithm from DeepMind, winner of the CASP14 challenge (2020) to predict distance and create dihedral maps using deep multiple sequence alignments as input. This algorithm does not directly use templates to place atomic coordinates but can use templates as part of the map building process.
NovaFold AI-Multimer uses DeepMind’s AlphaFold-Multimer algorithm, a newer extension of AlphaFold 2 that was built to predict multiple chain protein complexes, also known as “multimers.” This algorithm was utilized as the foundational element in the three top competitors in the multimer category of CASP15 (2022).
NovaFold utilizes the award-winning I-TASSER protein structure prediction algorithm, which combines threading and ab initio folding technologies. NovaFold finds more distant relationships than sequence similarity alone, improving the quality and accuracy of the predicted structure.
Protein structure prediction in 4 simple steps
Step 1
Select your protein sequence or region of interest
Step 2
Run the prediction using NovaFold, NovaFold AI, or NovaFold AI-Multimer
Step 3
View model and analyze structural data, predicted binding sites, and protein function
Step 4
Open or align predicted models for further analysis using Protean 3D
Resources
Please see our resources below for more information on protein structure prediction.
Protein Structure Prediction with NovaFold AI & NovaFold AI-Multimer
Protein Structure Prediction with NovaFold AI-Multimer
Protein Structure Prediction with NovaFold AI
NovaFold AI Protein Structure Prediction Software
Use NovaFold AI to Predict a Protein Structure with a Cytosolic Domain
Use NovaFold to Predict a Protein Structure
Accuracy of NovaFold Protein Folding Predictions
Modeling GPCR Structures in NovaFold
Why Structure Prediction Matters
Tutorials
Watch one of our videos or check out our user guide to learn more about using NovaFold and NovaFold AI for 3D structure prediction.
3D Structure Prediction of a Protein Using NovaFold AI
DNASTAR’s NovaFold AI application uses the award-winning AlphaFold 2 algorithm to predict the 3D structure of a protein based on its sequence.
Lasergene Protein Overview
This video gives an overview of Lasergene Protein, consisting of Protean 3D, plus the optional services NovaFold, NovaFold Antibody and NovaDock.
NovaFold Structure Prediction Software
This video gives an overview of NovaFold, the solution for all of your structure prediction needs. NovaFold uses a unique hybrid approach of threading and ab initio folding, allowing you to create 3D atomic models of proteins of any size with the highest accuracy.
Analyzing NovaFold Protein Structure Prediction Results
In this video, learn how to use the NovaFold report and analysis views to analyze your protein structure prediction models, including how to examine predicted ligand binding sites, align experimental structures to the prediction model, analyze various statistics and more.
FAQs
What is the protein folding prediction method used?
NovaFold, NovaFold AI & NovaFold AI-Multimer and are all services that run through the Protean 3D interface. NovaFold uses the I-TASSER algorithm developed by Professor Yang Zhang of the…
NovaFold, NovaFold AI & NovaFold AI-Multimer and are all services that run through the Protean 3D interface.
NovaFold uses the I-TASSER algorithm developed by Professor Yang Zhang of the Department of Computational Medicine and Bioinformatics at the University of Michigan. This protein structure prediction algorithm utilizes a combination of “threading” and “ab initio folding” in predicting protein structure. Threading attempts to match portions of the query sequence to template sequences. The template sequences, and their experimentally solved structures, are part of the RCSB Protein Data Bank (PDB). Ab initio folding uses biophysical properties of the query sequence and simulations to determine the likely structure(s) of the protein.
NovaFold AI uses AlphaFold 2, an artificial intelligence-based algorithm developed by Google’s DeepMind, to predict distance and create dihedral maps using deep multiple sequence alignments as input. This algorithm does not directly use templates to place atomic coordinates but can use templates as part of the map building process.
NovaFold AI-Multimer uses AlphaFold-Multimer, another AI-based algorithm developed by Google’s DeepMind as an extension of AlphaFold 2. AlphaFold-Multimer is optimized to predict the structures of multi-chain protein complexes, also known as multimers.
How accurate are protein folding predictions?
The “gold standard” for determining the accuracy of protein structure prediction algorithms is the CASP (Critical Assessment of Protein Structure Prediction) challenge…
The “gold standard” for determining the accuracy of protein structure prediction algorithms is the CASP (Critical Assessment of Protein Structure Prediction) challenge, organized by the Protein Structure Prediction Center. Each CASP experiment is a biennial, world-wide evaluation of structure prediction methods, with approximately 100 participating laboratories entering one or more different algorithms into competition. Each algorithm must predict dozens of structures based on their protein sequences alone. The predictions are then mathematically compared to each structure as determined through x-ray crystallography.
In CASP competitions from 2006-2018, the I-TASSER (also called “Zhang-Server”) algorithm used by NovaFold was ranked:
CASP7 – 1st
CASP8 – 1st
CASP9 – 1st (tie)
CASP10 – 1st
CASP11 -1st
CASP12 – 3rd
CASP13 – 2nd
The CASP14 experiments in 2020 tested 146 algorithms. AlphaFold 2, the algorithm used in NovaFold AI, made its debut in this competition, and earned the first-place ranking. In fact, AlphaFold 2 was an astonishing 2.65x more accurate at predicting protein structures than the second-ranked algorithm. I-TASSER still put in a strong showing, coming in at 5th place.
AlphaFold-Multimer was the foundational algorithm used by the top three competitors in the multimer target portion of the CASP15 experiments in 2022.
How do I visualize the predicted structures?
Results of a completed NovaFold AI or NovaFold AI-Multimer prediction can be opened and viewed just like any other structure in Protean 3D. By contrast, NovaFold predictions…
Results of a completed NovaFold AI or NovaFold AI-Multimer prediction can be opened and viewed just like any other structure in Protean 3D. By contrast, NovaFold predictions are shown in a specialized Report view in Protean 3D. After running a NovaFold prediction, the Report view opens automatically. The Report uses interactive images and tables to show which templates were used in the prediction and the models that were the best match to the query. Scroll to the prediction model of interest and click the link “Open model in new document” to open it as a new Protean 3D document for further analysis.
What statistics are provided for analyzing 3D protein structure prediction results?
Structure predictions include statistics such as the template modeling score (Tm-score), root mean square deviation (RMSD), confidence score, cluster size, density score, and more. The Protean 3D User Guide provides information about each of these scores and how they are used to determine the “best fit” model predictions.
What file types can I import to predict structures?
Protean 3D supports the following protein file formats: .aa, .fap, .fas, .fasta, .gp, .gbk, .sbd, .pro.
Can I export the predicted structures for publication?
Yes. Start by opening the predicted structure in Protean 3D. For NovaFold only, you will need to scroll down the Report view for the prediction model of interest and click the link “Open model in new document.” You can then export an image of the predicted structure in .png, .jpg or .gif format using File > Export Image > Structure. To instead export the predicted structure as a .pdb or .cif file, use one of the File > Export Data subcommands.
Are there any differences between protein structure prediction in NovaFold compared to I-TASSER?
While NovaFold is largely based on I-TASSER, we’ve enhanced NovaFold protein structure predictions in several ways:
- NovaFold predictions include improved bond geometries and reduced atomic clashes, creating energetically…
While NovaFold is largely based on I-TASSER, we’ve enhanced NovaFold protein structure predictions in several ways:
-
- NovaFold predictions include improved bond geometries and reduced atomic clashes, creating energetically minimized structure models.
-
- NovaFold can also search custom template libraries for a more specialized approach to homology modeling – contact support@dnastar.com or request a webinar for information on building a custom library.
-
- NovaFold is fully integrated with Protean 3D, creating a seamless experience for visualizing and analyzing your predicted protein models.
All of these enhancements build on the foundational I-TASSER algorithms in NovaFold.
Citations
Structure/epitope analysis and IgE binding activities of three cyclophilin family proteins from Dermatophagoides pteronyssinus
Li, Y., Sun, X. & Yang, L (2023). Sci Rep 13, 13630. https://doi.org/10.1038/s41598-023-40720-6.
The role of Streptococcal cell-envelope proteases in bacterial evasion of the innate immune system
Sophie McKenna, Kristin Krohn Huse, Sean Giblin et al. (2022). J Innate Immun 4 April 2022; 14 (2): 69–88. https://doi.org/10.1159/000516956.
Kir7.1 disease mutant T153I within the inner pore affects K+ conduction
Katie M. Beverley, Pawan K. Shahi, Meha Kabra et al. (2022), American Journal of Physiology-Cell Physiology 323:1, C56-C68. https://doi.org/10.1152/ajpcell.00093.2022.
Both recombinant Bacillus subtilis expressing PCV2d Cap protein and PCV2d-VLPs can stimulate strong protective immune responses in mice
Zhang Y, Wu Y, Peng C et al. (2023). Heliyon Volume 9, Issue 12, E22941, Dec 2023. https://doi.org/10.1016/j.heliyon.2023.e22941.
Characterization of a virulence factor in Plasmodiophora brassicae, with molecular markers for identification
Sedaghatkish A, Gossen BD, McDonald MR (2023). PLoS ONE 18(9): e0289842. https://doi.org/10.1371/journal.pone.0289842.
Ligands exert biased activity to regulate sigma 1 receptor interactions with cationic TRPA1, TRPV1, and TRPM8 dhannels
Cortés-Montero E, Sánchez-Blázquez P, Onetti Y, Merlos M and Garzón J (2019) Front. Pharmacol. 10:634. https://doi.org/10.3389/fphar.2019.00634.
The axonal motor neuropathy-related HINT1 protein is a zinc- and calmodulin-regulated cysteine SUMO protease
Elsa Cortés-Montero, María Rodríguez-Muñoz, Pilar Sánchez-Blázquez, and Javier Garzón (2019). Antioxidants & Redox Signaling, Vol. 31, No. 7 Sep 2019.
A bifunctional cellulase–xylanase of a new Chryseobacterium strain isolated from the dung of a straw-fed cattle
Tan, Hao et al. (2018). Microbial Biotechnology 11(2), 381– 398. https://doi.org/10.1111/1751-7915.13034.