Research Interests
High-throughput technologies—such as DNA sequencing, gene and protein expression profiling, DNA copy number analysis, and single nucleotide polymorphism genotyping—produce large amounts of data and have created a need for new tools that can assist in extracting the significant biological information from these data sets. Bioinformatics and computational biology are new disciplines that develop methods for the storage, distribution, integration, and analysis of these large data sets. The Computational Biology laboratory at VARI uses mathematical and computer science approaches to analyze and integrate complex data sets with a goal of understanding how cancer cells differ from normal cells at the molecular level. In addition, members of the lab provide assistance in data analysis and other computational projects on a collaborative and/or fee-for-service basis.
In the past year, the laboratory has taken part in many collaborative projects to further the research efforts at VARI. We have contributed gene expression analysis to projects ranging from identifying mechanisms of oncogene transformation to identifying genes associated with drug resistance. In recent work led by the Laboratory of Chromosome Replication, we examined how the deregulation of genes involved in chromosome replication are associated with the development and progression of several types of cancer. We have worked closely with the Laboratory of Cancer Genetics in developing gene expression–based models for the diagnosis and prognosis of renal cell carcinoma. We are also part of a multi-lab project spearheaded by the Laboratory of Cancer and Developmental Cell Biology to identify and characterize genes associated with the development of hereditary hemangiosarcomas in canines. Our role in this project focuses on the integration of data from single nucleotide polymorphism, gene expression, and pathway modeling studies.
In addition to collaborative work, the lab has a particular interest in developing and applying computational models that use gene expression data to identify large chromosomal abnormalities in cancer cells. In humans, each cell contains a set of approximately 6 billion DNA bases that are packaged into 46 chromosomes. From these chromosomes, at least 20,000 different types of messenger RNAs (mRNAs) and hundreds of non-coding RNAs (ncRNAs) are produced. Structural changes in chromosomes, such as translocations, deletions, rearrangements, and amplifications, commonly occur in cancer cells and likely contribute to the development and progression of the disease through disruptions in RNA production. We are building computational tools that use RNA expression to both identify chromosomal abnormalities and identify which single RNA (or set of RNAs), when deregulated, contributes to tumor development. In recent work, these RNA-based models predicted that high-grade papillary renal cell carcinoma contained a chromosome 8q amplification associated with overexpression of the c-MYC gene and activation of the MYC transcriptional program. This prediction was subsequently confirmed using molecular and cell biology experiments, highlighting the potential of gene expression profiling data for building integrative computational models of tumor development and progression.
The use of RNA-based models has the potential to identify even more-subtle chromosomal changes, such as changes in chromosome conformation. Examination of gene expression data derived from a subtype of renal cancer, renal oncocytoma, revealed that the population of RNAs produced from chromosome 19 was significantly up-regulated relative to the RNAs produced in normal kidney cells. Although no structural abnormality on chromosome 19 was identified, a more detailed cytogenetic analysis of renal oncocytoma cells showed that the chromosome 19 homologues had become intertwined or “paired”. The pairing was associated with the changes in the amount of mRNA produced from this chromosome. We are currently working to determine if chromosome pairing is present in other types of tumor cells and to determine the role of the chromosomal state in tumor development and progression.
