DiANNA: a web server for disulfide connectivity prediction

By Fabrizio Ferrè, Peter Clote. Nucleic Acids Res., 2005

388 citations recorded by the source corpus.

Correctly predicting the disulfide bond topology in a protein is of crucial importance for the understanding of protein function and can be of great help for tertiary prediction methods. The web server outputs the disulfide connectivity prediction given input of a protein sequence. The following procedure is performed. First, PSIPRED is run to predict the protein's secondary structure, then PSIBLAST is run against the non-redundant SwissProt to obtain a multiple alignment of the input sequence. The predicted secondary structure and the profile arising from this alignment are used in the training phase of our neural network. Next, cysteine oxidation state is predicted, then each pair of cysteines in the protein sequence is assigned a likelihood of forming a disulfide bond—this is performed by means of a novel architecture (diresidue neural network). Finally, Rothberg's implementation of Gabow's maximum weighted matching algorithm is applied to diresidue neural network scores in order to produce the final connectivity prediction. Our novel neural network-based approach achieves results that are comparable and in some cases better than the current state-of-the-art methods.

Subfields: Computational Theory and Mathematics, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/f874a91e85eb660842ceee3b93fb06a421b8519e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

1. Alessio Ceroni
2. Andrea Passerini
3. Paolo Frasconi
4. Alessandro Vullo

In Molecular Biology

1. Alessio Ceroni
2. Andrea Passerini
3. Paolo Frasconi
4. Alessandro Vullo

Citations in the corpus

• DISULFIND: a disulfide bonding state and cysteine connectivity prediction server (2006, methodology): “Ferre` and Clote (13) proposed a feedforward neural-network architecture with hidden units associated with cysteine pairs and inputs encoding secondary structure; the method is behind the prediction server DiANNA (14).”