Profile-based string kernels for remote homology detection and motif extraction

By Rui Kuang, Eugene Ie, Ke Wang, Kai Wang, Mahira Siddiqi, Yoav Freund, Christina S. Leslie. 2004

230 citations recorded by the source corpus.

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

Discriminative Embeddings of Latent Variable Models for Structured Data (2016, positive): “We use the same 54 target families and the same training/test splits as in remote homology detection (Kuang et al., 2005).”
Biological Sequence Modeling with Convolutional Kernel Networks (2017, neutral): “Proﬁle-based methods [Kuang et al., 2005, Rangwala and Karypis, 2005] have shown very good performance on this task but suﬀer a few limitations as pointed out by [Hochreiter et al., 2007], including computation time and interpretability.”
Novel Machine Learning Methods for MHC Class I Binding Prediction (2010, methodology): “They have been successfully applied to various problems in computational biology, ranging from protein remote homology detection [10],to gene identification [16, 20], to drug design [8].”
A new learning paradigm: Learning using privileged information (2009, unknown): “The similarity between amino-acid sequences was computed using the profile-kernel (Kuang et al., 2005) and that between3D-structureswas computedusingMAMMOTH a program to compare 3D-structures (Ortiz, Strauss, & Olmea, 2002).”
Fast model-based protein homology detection without alignment (2007, methodology): “LSTM is applied to the PROSITE database (Bairoch, 1995), and the motifs extracted by LSTM are compared to the PROSITE motifs (Sigrist et al.,2002).”
Modelling input texts: from Tree Kernels to Deep Learning (2015, background): “Structural kernels have been successfully applied to ease the design of machine learning systems in diverse domains, ranging from bioinformatics [78,82,130] and data-mining [10, 30,139,162,173,181,184,191] to Natural Language Processing (NLP) [23,34,37,73,79,142].”
LocTree3 prediction of localization (2014, unknown): “LocTree2 uses advanced SVM profile kernels (16).”
Modeling the language of life – Deep Learning Protein Sequences (2019, methodology): “This concept resembles 24 that of using k-mers for database searches [33], clustering [74], motifs [75, 76], and prediction 25 methods [66, 70, 77-81].”
Semi-supervised Protein Classification Using Cluster Kernels (2003, methodology): “More recently, Kuang et al. (2004) introduced a semi-supervised profile-based string kernel for protein sequences.”
Transfer String Kernel for Cross-Context DNA-Protein Binding Prediction (2016, background): “(3) The profile kernel [17] uses the notion of similarity base d on a probabilistic model (e.”
Protein-ligand interaction prediction: an improved chemogenomics approach (2008, background): “For each family, the database provides a list of known compounds for each target.”