A Discriminative Framework for Detecting Remote Protein Homologies

By Tommi S. Jaakkola, Mark Diekhans, David Haussler. J. Comput. Biol., 2000

578 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/23edfcc4096b293306cc1ac7864f2e83ac3bb1b8

How researchers see this paper

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

In Molecular Biology

Citations in the corpus

The Spectrum Kernel: A String Kernel for SVM Protein Classification (2001, methodology): “These include the original experimental results from Jaakkola et al. for two methods: the SAM-T98 iterative HMM, and the Fisher-SVM method.”
Profile-based string kernels for remote homology detection and motif extraction (2004, methodology): “Among the approaches that have been presented are the Fisher-SVM method [12], which represents each protein sequence as a vector of Fisher scores extracted from a profile hidden Markov model (HMM) for a protein family, and kernels that extend the Fisher kernel method [20]; families of efficient…”
Linear-Time Computation of Similarity Measures for Sequential Data (2008, methodology): “A further alternative for comparison of sequences are kernels derived from generative probability models, such as the Fisher kernel (Jaakkola et al., 2000) and the TOP kernel (Tsuda et al., 2002).”
Neural methods for non-standard data (2004, background): “The Fisher kernel constitutes an early and very prominent approach within this line [41].”
Biological Sequence Modeling with Convolutional Kernel Networks (2017, unknown): “Biological sequences are typically embedded into a space of very large of infinite dimension and represented as a rich set of descriptors, based for instance on the Fisher score [14], k-mer spectrum up to some mismatches [24], or local alignments [32].”
A comprehensive review and comparison of different computational methods for protein remote homology detection (2018, methodology): “SVM-Fisher [76] is the first indirect kernel-based method for detecting remote protein homology, which maps all protein sequences into points in a Euclidean feature space based on the sequence similarity scores output by HMM alignment.”
Fast Kernels for String and Tree Matching (2002, methodology): “For a proof of concept we tested our approach on a remote homology detection problem 1 [9] using Stafford Noble's SVM package2 as the training algorithm.”
Combining Pairwise Sequence Similarity and Support Vector Machines for Detecting Remote Protein Evolutionary and Structural Relationships (2003, methodology): “The SVM implementation employs the optimization algorithm described by Jaakkola et al. (2000) For each family, the numbers of sequences in the positive and negative training and test sets are listed.”
Protein-ligand interaction prediction: an improved chemogenomics approach (2008, background): “For each family, the database provides a list of known compounds for each target.”
Learning Discriminative Fisher Kernels (2011, methodology): “…as a principled way to use the power of probabilistic generative models in kernel methods (Jaakkola & Haussler, 1998), and have since been successfully used in numerous applications, e.g., in protein homology detection (Jaakkola et al., 2000) and speaker recognition (Campbell et al., 2006).”
Support vector machines (2008, unknown): “In particular, the challenge of uncovering structure in DNA has influenced several extensions of the feature map in (18), for example by incorporating generative models (Jaakkola et al., 2000; Tsuda et al., 2002), inexact and position-dependent matching (Leslie et al.”