Informatics for Unveiling Hidden Genome Signatures

Abstract

With the increasing amount of available genome sequences, novel tools are needed for comprehensive analysis of species-specific sequence characteristics for a wide variety of genomes. We used an unsupervised neural network algorithm, a self-organizing map (SOM), to analyze di-, tri-, and tetranucleotide frequencies in a wide variety of prokaryotic and eukaryotic genomes. The SOM, which can cluster complex data efficiently, was shown to be an excellent tool for analyzing global characteristics of genome sequences and for revealing key combinations of oligonucleotides representing individual genomes. From analysis of 1-and 10-kb genomic sequences derived from 65 bacteria (a total of 170 Mb) and from 6 eukaryotes (460 Mb), clear species-specific separations of major portions of the sequences were obtained with the di-, tri-, and tetranucleotide SOMs. The unsupervised algorithm could recognize, in most 10-kb sequences, the species-specific characteristics (key combinations of oligonucleotide frequencies) that are signature features of each genome. We were able to classify DNA sequences within one and between many species into subgroups that corresponded generally to biological categories. Because the classification power is very high, the SOM is an efficient and fundamental bioinformatic strategy for extracting a wide range of genomic information from a vast amount of sequences.[Supplemental material is available online at www.genome.org.]In addition to protein-coding information, genome sequences contain a wealth of information of interest in many fields of biology, from molecular evolution to genome engineering. G+C% has been used as a fundamental characteristic of individual genomes, but the G+C% is apparently too simple a parameter to differentiate a wide variety of genomes of known sequences. Oligonucleotide frequency can be used to distinguish genomes, because oligonucleotide frequencies vary significantly among genomes; dinucleotide frequencies, for example, are shown to be genome signatures for both prokaryotes and eukaryotes (Nussinov 1984;Karlin et al. 1997;Karlin 1998;Gentles and Karlin 2001). Comprehensive analyses of oligonucleotide frequencies in a wide variety of genomes are thought to provide fundamental knowledge of individual genomes, namely, key combinations of oligonucleotides responsible for the biological properties of the different genomes and genome portions. We applied Kohonen's self-organizing map (SOM) to create graphical representations of oligonucleotide frequencies from which we could extract a wide range of genomic information. The unsupervised neural network algorithm is an effective tool for clustering and visualizing high-dimensional data; it converts complex nonlinear relations among high-dimensional data into simple geometric relations that can be viewed in two dimensions (Kohonen 1982(Kohonen , 1990Kohonen et al. 1996).We and others have used SOMs to characterize codon usage patterns of a wide variety of bacteria (Kanaya et al. 1998;Wang et al. 2001). We introduced a new feature ...

Novel organization of aromatic degradation pathway genes in a microbial community as revealed by metagenomic analysis

“…A tRNA gene search was carried out using the tRNAscan-SE program (Lowe and Eddy, 1997). Self-organization map (SOM) analysis (Abe et al, 2003) was performed using each 5-kb DNA window of inserts as a query. Pairwise nucleotide sequence comparisons were made and visualized using the GenomeMatcher program (Ohtsubo et al, 2008).…”

A robust omics-based approach for the identification of glucosinolate biosynthetic genes

“…Time-series data for the transcriptome and the metabolome of leaves and roots were obtained, and analyzed by batch-learning selforganizing mapping (BL-SOM), a sophisticated form of multivariate analysis (Abe et al 2003;Kanaya et al 2001). BL-SOM, along with other clustering algorithms such as k-means and hierarchical clustering, can be used for co-occurrence analysis of genes and metabolites.…”

Classification and regression tree (CART) analyses of genomic signatures reveal sets of tetramers that discriminate temperature optima of archaea and bacteria

“…Other techniques have been used to classify fragments of DNA according to nucleotide composition. These include machine learning methods such as the self-organizing map (SOM) (e.g., Kohonen 1990, Abe et al 2003, 2006 and the support-vector-machine (SVM) (e.g., Rigoutsos 2005, McHardy et al 2007). In other studies, e.g., Lin et al 2003, CART has allowed efficient use of large collections of classifying variables to identify nonlinear relationships among the classifiers, yielding a simple sequential set of classification rules.…”