Limits...
TranscriptomeBrowser: a powerful and flexible toolbox to explore productively the transcriptional landscape of the Gene Expression Omnibus database.

Lopez F, Textoris J, Bergon A, Didier G, Remy E, Granjeaud S, Imbert J, Nguyen C, Puthier D - PLoS ONE (2008)

Bottom Line: Over-representation of functional terms was found in a large proportion of these TS (84%).Finally, taking advantage of our large collection of transcriptional signatures, we constructed a comprehensive map that summarizes gene-gene co-regulations observed through all the experiments performed on HGU133A Affymetrix platform.We provide evidences that this map can extend our knowledge of cellular signaling pathways.

View Article: PubMed Central - PubMed

Affiliation: Inserm U928, TAGC, Parc Scientifique de Luminy, Marseille, France.

ABSTRACT

Background: As public microarray repositories are constantly growing, we are facing the challenge of designing strategies to provide productive access to the available data.

Methodology: We used a modified version of the Markov clustering algorithm to systematically extract clusters of co-regulated genes from hundreds of microarray datasets stored in the Gene Expression Omnibus database (n = 1,484). This approach led to the definition of 18,250 transcriptional signatures (TS) that were tested for functional enrichment using the DAVID knowledgebase. Over-representation of functional terms was found in a large proportion of these TS (84%). We developed a JAVA application, TBrowser that comes with an open plug-in architecture and whose interface implements a highly sophisticated search engine supporting several Boolean operators (http://tagc.univ-mrs.fr/tbrowser/). User can search and analyze TS containing a list of identifiers (gene symbols or AffyIDs) or associated with a set of functional terms.

Conclusions/significance: As proof of principle, TBrowser was used to define breast cancer cell specific genes and to detect chromosomal abnormalities in tumors. Finally, taking advantage of our large collection of transcriptional signatures, we constructed a comprehensive map that summarizes gene-gene co-regulations observed through all the experiments performed on HGU133A Affymetrix platform. We provide evidences that this map can extend our knowledge of cellular signaling pathways.

Show MeSH

Related in: MedlinePlus

The transcriptional MAP associated with GPL96 related experiments.(A) A low resolution image made of 22,215 probes from GPL96 platform as rows and 3,114 GPL96 specific TS as columns. Red color indicates the presence of a gene in the corresponding TS (default to black). (B) Zooms of the corresponding areas showing some immune system related meta-signatures. (C) Representative genes that fall into these clusters.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2602602&req=5

pone-0004001-g004: The transcriptional MAP associated with GPL96 related experiments.(A) A low resolution image made of 22,215 probes from GPL96 platform as rows and 3,114 GPL96 specific TS as columns. Red color indicates the presence of a gene in the corresponding TS (default to black). (B) Zooms of the corresponding areas showing some immune system related meta-signatures. (C) Representative genes that fall into these clusters.

Mentions: The paradigm that genes from a TS share functional relationships is now widely accepted and constitutes the basis of transcriptome analysis [17]. However, each of these TS is rather associated to multiple underlying pathways whose components and limits are unclear. Our difficulty in depicting comprehensive maps for pathways is illustrated by existing discrepancies, for instance, between those proposed by BioCarta, KEGG and GeneMAPP. We reasoned that the more frequently two genes fall in the same TS, the more likely these genes belong to the same core functional network. To test this hypothesis, we produced a Boolean matrix with 22,215 probes from GPL96 platform as rows and 3,114 GPL96 specific TS as columns (only TS containing 30 to 1500 probes were included). This matrix was filled with zero and elements were set to 1 if a given gene was observed in the corresponding TS. Hierarchical clustering with uncentered Pearson's correlation coefficient was used to reveal genes frequently associated to the same TS. Given the order of the resulting matrix, it could not be visualized on a desktop computer using conventional software (i.e.; Treeview, MeV). We thus developed the TBMap plugin which allows one to visualize the map but also to superimpose a user-defined or a KEGG-related gene list. As expected, most of the clusters where obviously enriched in genes involved in similar biological processes (Protein biosynthesis/Ribosome function, oxidative phosphorylation, cell cycle, fatty acid metabolism, valine leucine and isoleucine degradation, extracellular matrix, breast cancer cells, structural constituent of muscles, neuronal processes, etc.). This was particularly clear when KEGG pathway informations were superimposed (see Figure S6). The Figure 4 presents some of the clusters that were identified as related to immune system functions. We could find a signature defining T cells that contained numerous cell-surface markers (e.g. TCA@, CD2, CD3G, CD6, IL2RB, IL2RG, IL7R, IL21R and ICOS), signaling genes (ZAP70, LAT, LCK, ITK) and cytotoxicity-related genes (GZMA, GZMB, GZMH, GZMK and PRF1). Concerning B-cells, three clusters were observed. A large signature contains mature B-cell markers (CD19, CD22, CD72 and CD79B) and transcription factors important in B-cell development such as PAX5 and TCL1A. A second signature contains POU2AF1/OBF-1, together with its described targets: genes coding for immunoglobulin (IGHG1, IGHG3, IGHA1, IGHM, IGJ, IGKC and IGL) and the B-cell maturation factor, TNFRSF17/BCMA [18], [19]. The third B-cell signature contains cell surface markers found in immature B-cells (CD24, VPREB1, IGLL1/CD179B and CR2/CD21) in addition to transcription factors known to play a crucial role during early B-cell development (TCF3, SPIB and CUTL1). The NK signature contains eight genes of the Killer cell immunoglobulin-like receptors (KIR) family, 3 genes of the killer cell lectin-like receptor family in addition to other markers whose expression has been reported on the surface of NK cells (CD160, CD244/2B4 and CD226) [20], [21], [22]. It also contains TBX21/T-bet together with IL18R1, IL18RAP, IL12RB2 and IFNG. Importantly, the IL12/IL18 combination has been shown to be potent inducers of both TBX21/T-bet and IFNG in NK cells[23], [24]. In addition to MHC-Class I, MHC-Class II and macrophage related signatures, two pathways related to immune function are presented in Figure 4. The AP1 pathway is made of the prototypical immediate early genes and contains numerous transcription factors (EGR1, EGR2, FOS, FOSB, IER2, JUN, JUNB, KLF6, KLF4, KLF10, ATF3, BTG2 and BTG3) whose complex interplay has been reported earlier. Finally, a NFKB signature was also observed which, again, contains prototypical regulators (NFKIA, NFKIE, RELB, BCL-3 and MAP3K8/TPL2) and known targets (CCL20, CXCL3, IL1B, IL8 and SOD2). Altogether, these results underline the high relevance of the signatures obtained using this compilation of TS derived from GPL96 related GEO experiments.


TranscriptomeBrowser: a powerful and flexible toolbox to explore productively the transcriptional landscape of the Gene Expression Omnibus database.

Lopez F, Textoris J, Bergon A, Didier G, Remy E, Granjeaud S, Imbert J, Nguyen C, Puthier D - PLoS ONE (2008)

The transcriptional MAP associated with GPL96 related experiments.(A) A low resolution image made of 22,215 probes from GPL96 platform as rows and 3,114 GPL96 specific TS as columns. Red color indicates the presence of a gene in the corresponding TS (default to black). (B) Zooms of the corresponding areas showing some immune system related meta-signatures. (C) Representative genes that fall into these clusters.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2602602&req=5

pone-0004001-g004: The transcriptional MAP associated with GPL96 related experiments.(A) A low resolution image made of 22,215 probes from GPL96 platform as rows and 3,114 GPL96 specific TS as columns. Red color indicates the presence of a gene in the corresponding TS (default to black). (B) Zooms of the corresponding areas showing some immune system related meta-signatures. (C) Representative genes that fall into these clusters.
Mentions: The paradigm that genes from a TS share functional relationships is now widely accepted and constitutes the basis of transcriptome analysis [17]. However, each of these TS is rather associated to multiple underlying pathways whose components and limits are unclear. Our difficulty in depicting comprehensive maps for pathways is illustrated by existing discrepancies, for instance, between those proposed by BioCarta, KEGG and GeneMAPP. We reasoned that the more frequently two genes fall in the same TS, the more likely these genes belong to the same core functional network. To test this hypothesis, we produced a Boolean matrix with 22,215 probes from GPL96 platform as rows and 3,114 GPL96 specific TS as columns (only TS containing 30 to 1500 probes were included). This matrix was filled with zero and elements were set to 1 if a given gene was observed in the corresponding TS. Hierarchical clustering with uncentered Pearson's correlation coefficient was used to reveal genes frequently associated to the same TS. Given the order of the resulting matrix, it could not be visualized on a desktop computer using conventional software (i.e.; Treeview, MeV). We thus developed the TBMap plugin which allows one to visualize the map but also to superimpose a user-defined or a KEGG-related gene list. As expected, most of the clusters where obviously enriched in genes involved in similar biological processes (Protein biosynthesis/Ribosome function, oxidative phosphorylation, cell cycle, fatty acid metabolism, valine leucine and isoleucine degradation, extracellular matrix, breast cancer cells, structural constituent of muscles, neuronal processes, etc.). This was particularly clear when KEGG pathway informations were superimposed (see Figure S6). The Figure 4 presents some of the clusters that were identified as related to immune system functions. We could find a signature defining T cells that contained numerous cell-surface markers (e.g. TCA@, CD2, CD3G, CD6, IL2RB, IL2RG, IL7R, IL21R and ICOS), signaling genes (ZAP70, LAT, LCK, ITK) and cytotoxicity-related genes (GZMA, GZMB, GZMH, GZMK and PRF1). Concerning B-cells, three clusters were observed. A large signature contains mature B-cell markers (CD19, CD22, CD72 and CD79B) and transcription factors important in B-cell development such as PAX5 and TCL1A. A second signature contains POU2AF1/OBF-1, together with its described targets: genes coding for immunoglobulin (IGHG1, IGHG3, IGHA1, IGHM, IGJ, IGKC and IGL) and the B-cell maturation factor, TNFRSF17/BCMA [18], [19]. The third B-cell signature contains cell surface markers found in immature B-cells (CD24, VPREB1, IGLL1/CD179B and CR2/CD21) in addition to transcription factors known to play a crucial role during early B-cell development (TCF3, SPIB and CUTL1). The NK signature contains eight genes of the Killer cell immunoglobulin-like receptors (KIR) family, 3 genes of the killer cell lectin-like receptor family in addition to other markers whose expression has been reported on the surface of NK cells (CD160, CD244/2B4 and CD226) [20], [21], [22]. It also contains TBX21/T-bet together with IL18R1, IL18RAP, IL12RB2 and IFNG. Importantly, the IL12/IL18 combination has been shown to be potent inducers of both TBX21/T-bet and IFNG in NK cells[23], [24]. In addition to MHC-Class I, MHC-Class II and macrophage related signatures, two pathways related to immune function are presented in Figure 4. The AP1 pathway is made of the prototypical immediate early genes and contains numerous transcription factors (EGR1, EGR2, FOS, FOSB, IER2, JUN, JUNB, KLF6, KLF4, KLF10, ATF3, BTG2 and BTG3) whose complex interplay has been reported earlier. Finally, a NFKB signature was also observed which, again, contains prototypical regulators (NFKIA, NFKIE, RELB, BCL-3 and MAP3K8/TPL2) and known targets (CCL20, CXCL3, IL1B, IL8 and SOD2). Altogether, these results underline the high relevance of the signatures obtained using this compilation of TS derived from GPL96 related GEO experiments.

Bottom Line: Over-representation of functional terms was found in a large proportion of these TS (84%).Finally, taking advantage of our large collection of transcriptional signatures, we constructed a comprehensive map that summarizes gene-gene co-regulations observed through all the experiments performed on HGU133A Affymetrix platform.We provide evidences that this map can extend our knowledge of cellular signaling pathways.

View Article: PubMed Central - PubMed

Affiliation: Inserm U928, TAGC, Parc Scientifique de Luminy, Marseille, France.

ABSTRACT

Background: As public microarray repositories are constantly growing, we are facing the challenge of designing strategies to provide productive access to the available data.

Methodology: We used a modified version of the Markov clustering algorithm to systematically extract clusters of co-regulated genes from hundreds of microarray datasets stored in the Gene Expression Omnibus database (n = 1,484). This approach led to the definition of 18,250 transcriptional signatures (TS) that were tested for functional enrichment using the DAVID knowledgebase. Over-representation of functional terms was found in a large proportion of these TS (84%). We developed a JAVA application, TBrowser that comes with an open plug-in architecture and whose interface implements a highly sophisticated search engine supporting several Boolean operators (http://tagc.univ-mrs.fr/tbrowser/). User can search and analyze TS containing a list of identifiers (gene symbols or AffyIDs) or associated with a set of functional terms.

Conclusions/significance: As proof of principle, TBrowser was used to define breast cancer cell specific genes and to detect chromosomal abnormalities in tumors. Finally, taking advantage of our large collection of transcriptional signatures, we constructed a comprehensive map that summarizes gene-gene co-regulations observed through all the experiments performed on HGU133A Affymetrix platform. We provide evidences that this map can extend our knowledge of cellular signaling pathways.

Show MeSH
Related in: MedlinePlus