SeqMINER

Application data
Created by	Tao Ye, Arnaud Krebs
Biological application domain(s)	ChIP-seq
Technology	Any
Created at	IGBMC (Strasbourg)
Maintained?	Yes
Input format(s)	BED, SAM, BAM
Output format(s)	BED, png
Programming language(s)	Java
Licence	GPLv3
Operating system(s)	platform-independent

Summary: seqMINER is an integrated portable ChIP-seq data interpretation platform with optimized performances for efficient handling of multiple genomewide datasets. seqMINER allows comparison and integration of multiple ChIP-seq datasets and extraction of qualitative as well as quantitative information. seqMINER can handle the biological complexity of most experimental situations and proposes supervised methods to the user in data categorization according to the analysed features. In addition, through multiple graphical representations, seqMINER allows visualisation and modelling of general as well as specific patterns in a given dataset. Moreover, seqMINER proposes a module to quantitatively analyse correlations and differences between datasets.

"Error: no local variable "counter" was set." is not a number.

Correlative integration of multiple datasets. Most genomic studies aim to define the function of a particular regulatory factor by understanding globally how it affects other co-occuring events in a regulatory circuit (i.e. chromatin modifications, binding of other factors) and the consequences on outputs of the regulated system (i.e gene expression). Thus we designed seqMINER to allow the integration of multiple ChIP-seq datasets in a quick and user friendly manner. seqMINER uses a list of BED formatted genomic coordinates (i.e. binding sites of a particular factor, set of genes, set of promoters etc.) as a reference for investigating information in other genomic datasets. Three stages can be distinguished in the analysis process (Figure 1). First, in the data collection module, seqMINER collects the read density from a reference dataset over a user-defined window around a set of coordinates and then calculates the read density in the same window in one or multiple other data sets. Second, in the clustering module, seqMINER, uses a supervised clustering procedure (k-means) to organise the identified loci presenting similar read densities within the specified window. Third, in the visualization module, seqMINER allows at glance visualisation of the entire output dataset through various graphical representations. seqMINER proposes two related complementary tag (read) density based methods to analyse the signal enrichment status in multiple tracks. The first is a qualitative method that defines general patterns and functional sub-groups in the dataset: densities over a window around the reference coordinates can be calculated at different resolutions in multiple tracks (Figure 2A). The created matrix can be organized by supervised k-means clustering to isolate groups of loci having similar features as developed by (Heintzman et al. 2009; Heintzman et al. 2007). At this stage, clusters can optionally be reorganized manually according to the biological significance. Visualisation of the whole dataset is in this case achieved through heatmaps. This method allows easy visualisation of signal distribution over multiple loci and identifies general patterns over the dataset, which can be plotted as average profiles. Information on signal distribution, that can be an important biological feature, is conserved (i.e. broad, sharp enrichment peaks). However, visualisation of quantitative phenomenon is more complex. The second method allows extraction of quantitative information. Raw tag counts as well as normalised enrichment over a control track can be calculated. The created matrix allows easy plotting of quantitative information and interpretation for one to one comparisons. Moreover, the method produces numerical data that are necessary to integrate sequencing data in larger mathematical models of a particular system. All visual features as well as the list of loci can be exported for further analysis with other methodologies (i.e. gene annotation (Krebs et al. 2008), ontology (Dennis et al. 2003)). Moreover, output data can be used in new clustering rounds using different sets of data or analysis parameters. This possibility facilitates multiple iterative steps of analysis inherent to the genomic data analysis.

Links

• SeqMINER Homepage [ edit link ]

• Add a Link

References

1. . 2010. Nucleic Acids Research

To add a reference for SeqMINER, enter the PubMed ID in the field below and click 'Add'.

 

[ edit box ]

Search for "SeqMINER" in the SEQanswers forum / BioStar or:

Web Search	Wiki Sites	Scientific
Google. Bioinformatics journals Clusty	WikiPedia MetaBase BioPedia OpenWetWare bifx wiki	PubMed Europe PMC ScientificCommons.org