methods and algorithms implemented in our analysis applications

A brief selection of the most commonly applied methods that we have in use. This list is by no means exhaustive, and if you are interested in something that you don't see on the list, please get in touch.
An illustration of how various methods fit into a systems-scale analysis can be found here.

Quick links: sequence assembly; - gene finding; - genome/ gene annotation; - transcription factor, signal sequence and general pattern finding; - motif discovery; - motif finding; - gene\ protein family discovery; - structural predictions; - alignment; - phylogenetic reconstruction; - microarray data analyses

sequence assembly (all with quality clipping)

• hashing and Smith-Waterman algorithm

• suffix tree algorithm

• dynamic programming

• hash table searching and sorting

gene finding

• supervised learning leading to HMM-based genefinding

• splice site modeling and interpolated markov models

• pair-HMM based

• generalised hidden markov model frameworks

• fourier transform

[back to top] [back to workflow]

genome/ gene annotation

• pair-HMM based

• dynamic programming local alignment

• EST to genome alignment

• simple similarity search

• comparative genomics

[back to top] [back to workflow]

transcription factor, signal sequence and general pattern finding

• general HMM based methods

• knowledge based comparison methods

• custom database search methods

[back to top] [back to workflow]

motif discovery

• general-purpose Gibbs sampling

• expectation Maximisation

• alignment-based methods

• pattern branching

[back to top] [back to workflow]

motif finding

• position specific scoring matrix (PSSM)

• postion specific probability matrix (PSPM)

• hidden markov models (HMM)

• prosite annotation custom fuzzy searching

• repeat finding

[back to top] [back to workflow]

gene/ protein family discovery

• markov cluster (MCL) algorithm

• adjacency matrix methods based on dynamic programming alignment

• adjacency matrix methods based on BLAST

• Fruchterman-Rheingold self organisation

[back to top] [back to workflow]

structural predictions

• neural network systems

• linear statistical methods

• information theory methods

[back to top] [back to workflow]

alignment

• segment to segment local alignment (diagonals)

• needleman-wunsch

• smith-waterman

• hash/k-tuple based alignments

• full dynamic programming algorithms

[back to top] [back to workflow]

phylogenetic reconstruction

• maximum likelihood

• quartet puzzling with maximum likelihood

• maximum parsimony

• distance matrix methods (NOT RECOMMENDED)

[back to top] [back to workflow]

microarray data analyses

• normalisation: RMA, PLIER, MAS 5, linear regression, iterative log

• statistical analyses: SAM, ANOVA, HCL, DAM, TTEST, and many others.

[back to top] [back to workflow]

[home] [services_products] [contacts]

Copyright 2004 - 2005 Janssen Genomics  bioinformatics   http://janssen-genomics.com