Publications describing science behind PubGene

A literature network of human genes for high-throughput analysis of gene expression.
Jenssen et al., Nature Genetics 2001 28 (21-28) PubMed

Biology's name game (News Feature)
Pearson, Nature 2001 411631-2 PubMed

Mining the Bibliome (News Feature)
Alfred, Nature Reviews, Genetics 2001 2 (401) PDF 2.7M

Mining the Bibliome (News Feature)
Searls, Pharmacogenomics J. 2001 1 (88-89) PubMed

Mining the bibliome: searching for a needle in a haystack? New computing tools are needed to effectively scan the growing amount of scientific literature for useful information
Grivell, EMBO reports 2002 3 PubMed

Media coverage

• Computer Scientist, New Scientist, 30 April 2001
• Genetic Gold Mine, InformationWeek

PubGene applied in research

Connectivity can be used to identify key genes in DNA microarray data: a study based on gene expression in nasal polyps before and after treatment with glucocorticoids.
Benson et al. Acta Otolaryngol 2007 13 (1-6) PubMed
**

Combining evidence, biomedical literature and statistical dependence: new insights for functional annotation of gene sets.
Aubry et al., 2006 May 4;7:241. BMC Bioinformatics 2006 7 (241) PubMed
**

Associations Between Gene Expression in Breast Cancer and Patient Survival
Jenssen et al., Hum Genet 2002 111 (411-420) PubMed
** PubGene was used to identify groups of genes whose expression patterns were associated with survival in patients with breast cancer from a 8024-gene microarray set. Networks of genes that were tightly linked in the literature and at the same time contributed to low P-values were identified using PubGene’s Expression Analysis module. This approach confirmed the relevance of proteins that were associated with poor survival prognosis for breast cancer through contribution of their subunits as well as the relevance of proteins that contributed to poor survival being a part of a large pathway. Hence, such analysis may have a prognostic value.

Gene Expression Levels in Different Stages of Progression in Oral Squamous Cell Carcinoma
Kuo et al., Proc AMIA Symp 2002 (415-419) PubMed
** PubGene was used to determine the literature co-occurrences between 199 known genes whose expression was significantly up-regulated in samples taken from patients with oral squamous cell carcinoma, one of the most common worldwide types of cancer. PubGene’s Pathway Mapping module was used to link MeSH and Gene Ontology terms to the 199-gene list. It was confirmed that 36% of the genes from the list have been previously described in the literature as associated with that particular form of cancer, and an additional 42% have been associated with other types of cancer.

Microarray Analysis of Gene Expression in the Aging Human Retina
Yoshida et al., Inv. Opht. Vis. Sci., 2002 43 (2554-2560) PubMed
** PubGene was used to find gene associations to identify potential biological relationships among similarly upregulated genes in gene expression profiles of young and elderly human retinas. This was to identify candidate genes for aging associated retinal diseases. With the input of selected genes from microarray slides containing 2400 genes, PubGene made a citation-based gene network. In the elderly-dominant network, interleukin-1a (IL1A), known to be associated with inflammation and immunity was centered. The young-dominant network centered the neuropepide somatostatin, SST. This result was consistent with the knowledge that aging is associated with inflammatory response and gave somatostatin a putative more important role of retinal function.

Differential gene expression in premalignant human epidermis revealed by cluster analysis of serial analysis of gene expression (SAGE) libraries
van Ruissen F et al., FASEB J. 2002 2 (246-248)PubMed
** Serial Analysis of Gene Expression (SAGE) was used to perform quantitative analysis of gene expression on tissue from premalignant epidermal tissue (actinic keratosis), normal human epidermis and cultured keratinocytes. A total of 23599 different transcripts were identified. Cluster analysis was applied to the 300 transcripts that had the highest difference in expression between the sick and the healthy tissue. The analysis revealed two gene clusters where genes were up-regulated in the tumor cells compared to the non-tumor cells. PubGene's literature network tool was used to analyze the 17 known genes (those that had GeneBank accession number) from the 300 transcripts. Eight of these known genes made up a literature co-occurrence network, and gave hypothesis about genes that might be related by their function or expression patterns, providing insight for further analysis.

Gene Expression After Vaccination of Mice with Formulations of Diphtheria Toxoid or Tetanus Toxoid and Different Adjuvants: Identification of Shared and Vaccine-specific Genes in Spleen Lymphocytes
Regnstrom et al., Vaccine 2003 21 (2307-2317) PubMed
** PubGene was used to validate the significance of the shared genes for the immune response. 61 such genes were in this article subjected to the PubGene database and a literature network of 26 genes (34%) was received. The highest score was obtained for the proliferating cell nuclear antigen (PCNA). The PCNA protein is reported to be involved in DNA synthesis, DNA replication, DNA repair and cell cycle progression and is considered to be a novel actor in immune response. When submitting the list of the 26 genes to PubGene's MeSH Map tool, the returned term associations for these genes agreed well with immune response functions, reflecting the activation of T-lymphocytes. These genes might, therefore, be used as markers for autoimmune disease in the future.

Microarray Analysis of Tumor Necrosis Factor a Induced Gene Expression in U373 Human Glioblastoma Cells
Schwamborn et al., BMC Genomics 2003 4 (46) PubMed
** PubGene was queried with the list of 200 genes chosen from 7500 genes on the cDNA microarray used to detect tumor necrosis factor a (TNF) response in astrocytoma cells U373 based on 3-fold up- or downregulation. These cells have been widely used as a model for inflammatory cytokine actions in the nervous system. PubGene’s literature association network made, revealed that most of these genes showed no previous association with TNF. This suggests that there have been identified a high degree of new or novel genes induced by TNF. Co-regulation for many functional groups such as proteasome and ribosomal proteins were detected.

PEI - a Potent, but Not Harmless, Mucosal Immuno-stimulator of Mixed T-helper Cell Response and FasL-mediated Cell Death in Mice
Regnstrom et al., Gene Therapy 2003 10 (1575-1583) PubMed
** Polyetyleneimine (PEI) is a cationic polymer and is a very effective gene (DNA) delivery system to both cells in culture and in vivo. PEI is toxic and little is known about what types of immune responses it evokes. PubGene was used to make a literature gene association network of all immune response genes up-regulated in a sample of cells and PEI without the antigen-encoding DNA (PEI-). This was done to understand the relation between these genes. The genes from the achieved network, was submitted to the MeSH tool. The associations returned, supported the hypothesis that PEI- has important immuno-stimulatory abilities. Gene lists from clusters of co-regulated genes were subjected into PubGene’s subset network tool. The 73 of 139 genes participating in all the sub-networks were submitted to the MeSH tool. The returning categories suggested a well-established link to Alzheimer Disease, but also involvement in cellular processes such as cell cycle regulation, oncogenesis and differentiation. This indicates that free PEI might have a toxic character.

First comprehensive mapping of cartilage transcripts to the human genome
Yager et al., Genomics 2004 83 (772-789)PubMed
** To elucidate the transcriptional events underlying cartilage development and dysfunction a complete set of genes expressed in human cartilage was characterized. PubGene tools was used for literature co-citation analysis of two gene sets, a set 24 genes that were implicated in HDAC-mediated chromatin repression in cartilage and a set of 11 genes that were implicated in Swi/SNF-mediated chromatin activation in cartilage. PubGene literature co-citation analysis suggests that 13 of the 24 genes included in the first set and the 11 genes included in the second set may be organized into functional networks. PubGene analysis has revealed that PCAF was present in both literature networks, suggesting that this gene might provide a link between a general HDAC-dependent chromatin-repression pathway and a general Swi/SNF dependent chromatin-activation pathway in cartilage. The fact that mRNAs specifying these network were found in cartilage provides experimental evidence for the existence of both a general chromatin repression and a general chromatin activation mechanisms.

Transcriptome variations in human CaCo-2 cells: a model for enterocyte differentiation and its link to iron absorption
Bédrine-Ferran et al., Genomics 2004 84 (524-535)PubMed
** The aim of the study was to identify genes or gene networks involved in regulation of iron absorption. CaCo-2 cell was used as a model to characterize genes which expression varies during the differentiation by means of a transcriptional approach based on microarray experiments. Experiments led to the identification of 80 down-regulated, 50 up-regulated, and 56 invariant genes. PubGene tools were queried with the four lists of gene symbols corresponding to identified clusters of invariant, up, down and variant (up and down-regulated) genes. For each cluster annotations were retrieved with respect to molecular functions and biological processes. Ontological annotations highlight the decrease in the expression of cell cycle and DNA metabolism -associated genes as well as transcription, protein metabolism, signal transduction, and nucleocytoplasmic transport - associated genes. Likewise, genes which expression was increased appeared to be linked to cell adhesion, lipid and xenobiotic metabolism, iron transport and homeostasis, and immune response.