Prediction of protein stability changes for single‐site mutations using support vector machines
- Jianlin Cheng, Arlo Z. Randall, P. Baldi
- 21 December 2005
Computer Science, Biology
Proteins: Structure, Function, and Bioinformatics
The method can accurately predict protein stability changes using primary sequence information only, it is applicable to many situations where the tertiary structure is unknown, overcoming a major limitation of previous methods which require tertiary information.
SCRATCH: a protein structure and structural feature prediction server
- Jianlin Cheng, Arlo Z. Randall, Michael J. Sweredoski, P. Baldi
- 27 June 2005
Nucleic Acids Res.
SCRATCH is a server for predicting protein tertiary structure and structural features. The SCRATCH software suite includes predictors for secondary structure, relative solvent accessibility,…
A large-scale evaluation of computational protein function prediction
- P. Radivojac, W. Clark, I. Friedberg
- 27 January 2013
Today's best protein function prediction algorithms substantially outperform widely used first-generation methods, with large gains on all types of targets, and there is considerable need for improvement of currently available tools.
Improved residue contact prediction using support vector machines and a large feature set
- Jianlin Cheng, P. Baldi
- 2 April 2007
A new contact map predictor (SVMcon) that uses support vector machines to predict medium- and long-range contacts and can be modularly incorporated into a structure prediction pipeline.
Three-stage prediction of protein ?-sheets by neural networks, alignments and graph algorithms
- Jianlin Cheng, P. Baldi
Intelligent Systems in Molecular Biology
A modular approach to the problem of predicting/assembling protein beta-sheets in a chain by integrating both local and global constraints in three steps is proposed and yields significant improvements over previous methods.
An expanded evaluation of protein function prediction methods shows an improvement in accuracy
- Yuxiang Jiang, T. Oron, P. Radivojac
- 3 January 2016
The second critical assessment of functional annotation (CAFA) conducted, a timed challenge to assess computational methods that automatically assign protein function, revealed that the definition of top-performing algorithms is ontology specific, that different performance metrics can be used to probe the nature of accurate predictions, and the relative diversity of predictions in the biological process and human phenotype ontologies.
Accurate Prediction of Protein Disordered Regions by Mining Protein Structure Data
- Jianlin Cheng, Michael J. Sweredoski, P. Baldi
- 1 November 2005
Computer Science, Chemistry
Data mining and knowledge discovery
An ab initio predictor of disordered regions called DISpro is developed which uses evolutionary information in the form of profiles, predicted secondary structure and relative solvent accessibility, and ensembles of 1D-recursive neural networks.
A neural network approach to ordinal regression
- Jianlin Cheng, Zheng Wang, G. Pollastri
- 8 April 2007
IEEE World Congress on Computational Intelligence
An effective approach to adapt a traditional neural network to learn ordinal categories is described, a generalization of the perceptron method for ordinal regression, which outperforms a neural network classification method.
The CAFA challenge reports improved protein function prediction and new functional annotations for hundreds of genes through experimental screens
- Naihui Zhou, Yuxiang Jiang, I. Friedberg
- 29 May 2019
The third CAFA challenge, CAFA3, that featured an expanded analysis over the previous CAFA rounds, both in terms of volume of data analyzed and the types of analysis performed, concluded that while predictions of the molecular function and biological process annotations have slightly improved over time, those of the cellular component have not.
DeepSF: deep convolutional neural network for mapping protein sequences to folds
- Jie Hou, B. Adhikari, Jianlin Cheng
- 4 June 2017
Biology, Computer Science
A deep 1D‐convolution neural network (DeepSF) is developed to directly classify any protein sequence into one of 1195 known folds, which is useful for both fold recognition and the study of sequence‐structure relationship.