Protein secondary structure prediction based on position-specific scoring matrices.

By David C. Jones. 1999

Subfields: Information Extraction, 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 Information Extraction

In Computational Biology & Bioinformatics

Citations in the corpus

Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences (2019, methodology): “…we compare to (1) one-hot amino acid representations, which results in the prior that predicts the most likely label for each position’s amino acid, and (2) representations from family-level frequencies summarized in position-speciﬁc scoring matrices (“PSSM”s; Jones, 1999) for each protein.”
Kernel Belief Propagation (2011, methodology): “We first ran PSI-BLAST to generate the sequence profile (a 20 dimensional feature for each amino acid position), and then used this profile as features for predicting the folding structure (Jones, 1999).”
Protein Sequence Analysis Using the MPI Bioinformatics Toolkit (2020, methodology): “You could turn off secondary structure scoring if you suspect that the match between the query and template was scored highly because of a chance similarity of their PSIPRED-predicted or DSSP-determined secondary structures.”
Toward the estimation of the absolute quality of individual protein structure models (2010, methodology): “The same procedure is applied to calculate Z-scores for the agreement terms, i.e. for each structure in the two reference sets PSIPRED and ACCpro have been applied to model the background distribution of expected secondary structure and solvent accessibility prediction accuracy.”
Modeling aspects of the language of life through transfer-learning protein sequences (2019, background): “The concept was quickly taken up [4 – 8] and predictions improved even more with larger families increasing evolutionary information through diversity [9, 10].”
Deep Supervised and Convolutional Generative Stochastic Network for Protein Secondary Structure Prediction (2014, background): “Secondary structures are commonly classiﬁed to 8 states (Kabsch & Sander, 1983) or be further combined into 3 states (Singh, 2005).”
MetaPSICOV: combining coevolution methods for accurate prediction of contacts and long range hydrogen bonding in proteins (2014, methodology): “The secondary structure probabilities are derived from PSIPRED (Jones, 1999), and the solvent exposure from a derivative of PSIPRED called SOLVPRED.”
The HHpred interactive server for protein homology detection and structure prediction (2005, methodology): “The database HMMs have been precalculated and also contain secondary structure information, either predicted by PSIPRED, or assigned from 3D structure by DSSP (31).”
SWISS-MODEL: modelling protein tertiary and quaternary structure using evolutionary information (2014, background): “Sequence Pro-files, predicted secondary structure (SSpro (14), PSIPRED (15)), predicted solvent accessibility (ACCpro (14)), per-residue solvent accessibility, (NACCESS (S. Hubbard and J. Thornton)), secondary structure (DSSP (16)) are calculated and stored alongside the structure.”
A Deep Learning Network Approach to ab initio Protein Secondary Structure Prediction (2014, methodology): “and a logistic function of the PSSM score (used in PSI-PRED [18]):”