Showing papers by "Chad J. Creighton published in 2003"

Mining gene expression databases for association rules

Abstract: Motivation: Global gene expression profiling, both at the transcript level and at the protein level, can be a valuable tool in the understanding of genes, biological networks, and cellular states. As larger and larger gene expression data sets become available, data mining techniques can be applied to identify patterns of interest in the data. Association rules, used widely in the area of market basket analysis, can be applied to the analysis of expression data as well. Association rules can reveal biologically relevant associations between different genes or between environmental effects and gene expression. An association rule has the form LHS⇒RHS, where LHS and RHS are disjoint sets of items, the RHS set being likely to occur whenever the LHS set occurs. Items in gene expression data can include genes that are highly expressed or repressed, as well as relevant facts describing the cellular environment of the genes (e.g. the diagnosis of a tumor sample from which a profile was obtained). Results: We demonstrate an algorithm for efficiently mining association rules from gene expression data, using the data set from Hughes et al. (Cell, 102, 109‐ 126, 2000) of 300 expression profiles for yeast. Using the algorithm, we find numerous rules in the data. A cursory analysis of some of these rules reveals numerous associations between certain genes, many of which make sense biologically, others suggesting new hypotheses that may warrant further investigation. In a data set derived from the yeast data set, but with the expression values for each transcript randomly shifted with respect to the experiments, no rules were found, indicating that most all of the rules mined from the actual data set are not likely to have occurred by chance. Availability: An implementation of the algorithm using Microsoft SQL Server with Access 2000 is available at http://dot.ped.med.umich.edu:2000/pub/assoc rules/ assoc rules.zip. Our results from mining the yeast data set are available at http://dot.ped.med.umich.edu: 2000/pub/assoc rules/yeast results.zip. ∗ To whom correspondence should be addressed.

Profiling of pathway-specific changes in gene expression following growth of human cancer cell lines transplanted into mice

Abstract: Background Tumor cells cultured in vitro are widely used to investigate the molecular biology of cancers and to evaluate responses to drugs and other agents. The full extent to which gene expression in cancer cells is modulated by extrinsic factors and by the microenvironment in which the cancer cells reside remains to be determined. Two cancer cell lines (A549 lung adenocarcinoma and U118 glioblastoma) were transplanted subcutaneously into immunodeficient mice to form tumors. Global gene-expression profiles of the tumors were determined, based on analysis of expression of human genes, and compared with expression profiles of the cell lines grown in culture.

Making sense of microarray data to classify cancer

Abstract: Profiling gene expression using DNA arrays has had a tremendous impact on biomedical research. From a disease investigation point of view, applications of DNA microarrays include uncovering unsuspected associations between genes and specific clinical features of disease, resulting in novel, molecular-based disease classifications. Cancer is a case in point. Most published studies of cancers using DNA microarrays have either examined a pathologically homogeneous set of tumors to identify clinically relevant subtypes, for example, responders vs nonresponders, or pathologically distinct subtypes of cancer of the same lineage, for example, high-stage vs low-stage tumors to identify molecular correlates, or tumors of different lineages to identify molecular signatures for each lineage. A study of cutaneous T-cell lymphoma by Kari et al,1 published recently, typifies both what one hopes to gain from disease investigations using DNA microarrays and the limitations of such studies.

Expression of Matrix Metalloproteinase 9 (MMP-9/Gelatinase B) in Adenocarcinomas Strongly Correlated with Expression of Immune Response Genes

Abstract: Matrix metalloproteinases (MMPs) are endopeptidases considered to be important regulators of the microenvironment of cancer. While MMPs are traditionally associated with the extracellular matrix (ECM), here we provide new evidence from an analysis of gene expression profiles from human tumor tissue that MMP-9 (gelatinase B) is associated with elements of the immune system in a way analogous to the association of other MMPs, such as MMP-2 (gelatinase A), with components of the ECM. An analysis of three independent microarray datasets of lung adenocarcinomas from previous studies (Nat. Med. 8, 816-824 (2002); Proc. Natl. Acad. Sci. USA 98, 13790- 13795 (2001); Proc. Natl. Acad. Sci. USA, 98, 13784-13789 (2001)) showed that, in each dataset, out of the set of genes with significant correlations in mRNA expression to the expression of MMP9 (P < 0.005), a highly dispro- portionate number were found to be annotated in the Locuslink database as having a role in the anti-pathogen re- sponse. By comparison, out of the set of genes significantly correlated with the expression of MMP2, a highly dis- proportionate number were known components of the ECM. The same patterns observed in the lung data for both MMP2 and MMP9 were found as well in an additional published dataset of colon and ovarian adenocarcinomas (Am. J. Pathol. 159, 1231-1238 (2001)). The results of this study suggest a greater functional role for MMP-9 in the immune response to cancer than what may previously have been thought.