Evaluation and improvement of multiple sequence methods for protein secondary structure prediction

By James A. Cuff, Geoffrey J. Barton. 1999

Subfields: Computational Biology & Bioinformatics

How researchers see this paper

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

In Computational Biology & Bioinformatics

Citations in the corpus

MSA Transformer (2021, methodology): “We evaluate models on the CB513 test set (Cuff & Barton, 1999).”
Modeling aspects of the language of life through transfer-learning protein sequences (2019, background): “…from high-quality structures (< 3 Å) released after 2015 (and at most 30% PIDE to any protein of known structure in the PDB at the time); CB513 [70]: 513 non-redundant sequences compiled 20 years ago (511 after SIFTS mapping); CASP12 [71]: 21 proteins taken from the CASP12 free-modelling…”
RaptorX-Property: a web server for protein structure property prediction (2016, background): “For 3-state secondary structure, we follow the definition in (21), i.”
Transformer protein language models are unsupervised structure learners (2020, neutral): “…feed as input the couplings a mhij for each layer m , for each head h , and for j ∈ [ i − 5 , i + 5] , for a total of 7260 input features. we achieve Using just 100 of the 8678 training proteins, 79.3% accuracy on 3-class secondary structure prediction on the CB513 test set (Cuff & Barton, 1999).”
Secondary structure prediction with support vector machines (2003, methodology): “Consensus predictions have also been shown to improve upon individual methods (Cuff and Barton, 1999 ) and a system that makes use of the most accurate modern methods may provide further gains in accuracy.”
JPred : a consensus secondary structure prediction server (1999, methodology): “In our independent test this approach gave the Oxford University Press highest accuracy compared to all other combinations (Cuff and Barton, 1998).”
Cascaded multiple classifiers for secondary structure prediction (2000, unknown): “It has been argued recently that the way of decomposing these 8 states could have a dramatic effect on the accuracy of a method (Cuff and Barton, 1999).”
Protein secondary structure prediction based on position-specific scoring matrices. (1999, methodology): “JPRED server (Cuff & Barton, 1998), and so some improvement in the PHD results might be anticipated if a more recent implementation of the program was tested.”
Protein secondary structure prediction for a single-sequence using hidden semi-Markov models (2006, methodology): “The first rule is also known as the 'EHL' mapping [34,35], the second rule is the one used in PSIPRED [1] and earlier outlined by Rost and Sander [36], while, finally, the third rule is the common 'CK' mapping, which is the one used in BSPSS and other methods [17,37,38].”
Hidden-Unit Conditional Random Fields (2011, methodology): “In our protein secondary structure prediction experiments, we used the CB513 data set, which contains amino acid structures of 513 proteins (Cuff and Barton, 1999).”
A Unified Multitask Architecture for Predicting Local Protein Properties (2012, background): “The secondary structure prediction tasks show how multitask learning helps with small data sets: for the small, CB513 data set, the accuracy improves by 3.7%, whereas for the larger secondary structure prediction benchmark, accuracy increases by 1.4%.”
Comparing two K-category assignments by a K-category correlation coefficient (2004, methodology): “The protein prediction methods are: phd(Rost, 1996), samt99sec(Karplus et al., 1998), jpred (Cuff and Barton, 1999), psipred(Jones, 1999), prof king (Ouali and King, 1999), apssp2(Raghava, 2000), prospect(Xu and Xu, 2000), phdpsi(Przybylski and Rost, 2001), sspro2(Pollastri et al., 2002),…”
OXBench: A benchmark for evaluation of protein multiple sequence alignment accuracy (2003, unknown): “This finding is consistent with results for secondary structure prediction, where the addition of data rather than underlying algorithm appear to dominate improvements in accuracy [49,50].”