Plant expressed sequence tags databases: practical uses and the improvement of their searches using network module analysis

Abstract

Sequencing technology has been rapidly advancing. Giga-sequencers, which produce several gigabases of fragmented sequences per run, are attractive for decoding genomes and expressed sequence tags (ESTs). A variety of plant genomes and ESTs have been sequenced since the decoding of the genome of Arabidopsis thaliana, the model plant. ESTs are useful for functional analyses of genes and proteins and as biomarkers, which are used to identify particular tissues and conditions due to the specificity of their expression. Sequenced plant genomes and ESTs have been entered into public databases, where they are freely downloadable. Sequences representative of particular functions or structures have been collected from public databases to curate smaller databases useful for studying protein function. Here, we discuss the uses of the currently available plant EST datasets. We also demonstrate the use of network module analysis to perform more stable (or irrespective of the difference of performance in each analyzing PC) homology searches and to provide more information on molecular functions of plant ESTs and proteins.Key words: Database, expressed sequence tag (EST), homology search, network module analysis.Plant Biotechnology 28, 351-360 (2011) DOI: 10.5511/plantbiotechnology.11.0818a Abbreviations: BLAST, Basic Local Alignment Search Tool; EST, expressed sequence tags; GBFF, GenBank flatfile; GFF, General Feature Format; GPFF, GenPept flatfime. This article can be found at http://www.jspcmb.jp/ Published online September 25, 2011 perform more stable homology searches and to provide more information on molecular functions of ESTs, downsizing while maintaining the precision of homology search and also constructing local modules, in which sequences are highly homologous and thus belong to a group with a common feature, are useful. The PSI-BLAST algorithm and its derivatives (Altschul et al. 1997;Lee et al. 2009;Li et al. 2011) focus on sequenceto-sequence hits between multiple sequences. To evaluate relationships between multiple elements (e.g., gene or metabolite), network module analysis is a useful approach (Saito et al. 2008). Network module analyses have been applied to plant gene co-expression, in which a plant gene is related to other genes based on similar expression profiles (Aoki et al. 2007;Ficklin and Feltus 2011;Huber et al. 2007;Marino-Ramirez et al. 2009;Ogata et al. 2010;Winden et al. 2011). This approach allows a co-expression module, which includes coexpressed gene to be assigned to a particular biological process. By identifying homologies between sequences, network module analysis can be used to create, a homology network in which a sequence (node) is connected to other sequences on the basis of high homology. To perform such analysis for a homology network, we used our algorithm (Ogata et al. 2009) according to the following processes: 1) performing BLAST for any pairs of sequences, 2) calculating association indices between pairs as described in "Userfriendly tools for using plant EST a...

A database for retrieving information on SARS-CoV-2 S protein mutations based on correlation network analysis

“…The technique was also applied for correlation network analysis to detect local communities or network modules, in which nodes are tightly connected to each other. Correlation network analyses have been used to illustrate biological networks such as gene coexpression [ 17 , 18 ] and gene homology networks [ 19 – 21 ]. To depict the relationships between the mutations shared by sampled virus genomes, a correlation network approach is useful.…”

ConfeitoGUI: A toolkit for size-sensitive community detection from a correlation network

“…Communities that are identified from analyses of molecular biological datasets (e.g., those in plants [ 12 – 15 ]) are useful for characterizing functional and evolutionary traits, such as by co-expression of genes and co-accumulation of metabolites for functional traits and gene homology for evolutionary traits.…”

“…By varying the correlation index thresholds, communities of different size that represent different levels of functionality can be identified. On the basis of gene homology, Ogata and Suzuki [ 15 ] constructed a plant correlation network comprising 3,167 genes encoding cytochrome P450 (CYP) proteins. A correlation coefficient threshold of 0.5 in their network identified 217 CYP gene communities.…”