Biological applications areas
It is one of the characteristics of our research program that we expect
all CHIP investigators to focus on
applications areas in medicine or biology. For some, this focus arises
from a long-standing clinical or scientific interest but for all, it
serves as the proving ground on which quantitative or computational
approaches are tested for their performance and their relevance. Currently,
in addition to the application areas which have their own rubrics
on this page, other biomedical applications include: neuroscience, development,
diabetes, aging, transplantation biology, and cancer.
Across all these biological application areas, availability of tissue,
primarily human, in sufficient sample sizes has been a persistent problem.
This problem has been the focus of our NCI-funded Shared Pathology Informatics
Network project.
|
Public tools, data models and data exchange
We believe that investigations of the massive amount of information generated
by the human genome project could benefit from a higher level of integration
and from increased ease of access. We aim at developing
tools able to present
genetic data in a format that more closely matches the requirements of
the researchers who will use them. An example of this is SNPper,
a web-based application that we developed to assist researchers in designing
and performing large-scale association studies. By integrating various
databases of genes and Single Nucleotide Polymorphisms (SNPs), SNPper
provides the user with powerful search capabilities that make it possible
to retrieve sets of SNPs according to their position in the genome. In
addition, we are developing tools to automatically link together databases
containing related information, but coming from different sources (for
example: linking a gene symbol to its PubMed entry, its OMIM code, the
microarray labels that refer to it, clinical annotations, etc.). Like others,
we firmly believe that these tools are important but not an end unto
themselves. We are equally convinced that continued evolution of these
tools will depend on a critical mass of investigators with expertise in the
computational or statistical sciences and deep knowledge of the biomedical
sciences. The necessary expertise cannot come from biology alone (see this survey
of biomedical informatics expertise). |
Population Genomics
We are involved with several large epidemiological studies to understand
the genetic component of several diseases. In addition to implementing
the databases, we are developing novel
methods of finding robust correlations
between phenotype and genotype. |
Gene Expression in Inflammatory Myopathies
The goals here are: (1) to formulate
and validate hypotheses relevant to the pathogenesis of the inflammatory
myopathies through the use of DNA microarrays and measurement of large-scale
muscle tissue gene expression, (2) to characterize patterns of muscle
gene expression among distinct clinical subtypes of inflammatory myopathies
and correlate these patterns with clinical phenotypes, and (3) to explore
possible gene function for genes, cDNAs, and expressed sequence tags
(ESTs) of unknown function through computational techniques applied to
these expression profiles. |
Similarity Measures in DNA Microarray Dataset Analysis for Functional
Genomics
Several methodologies are available to explore functional relationships
among genes as inferred from DNA microarray expression analysis. The choice
of similarity measure in these varied techniques of functional genomic
clustering is the most significant determinant of the resulting hypothesized
relationships (e.g. ones based on dynamics, signal
coherence). Consideration
of the distinct mathematical properties of similarity measures provides
insight into their appropriate use in gene expression datasets as well
allowing for abstractions to other less intuitive similarity measures. |
Noise, Error and Reproducibility
Comparisons between expression measurements from repeated samples on duplicate
identical microarrays. Analysis of quanititation
algorithms in microarray array scanning software. Comparisons of expression measurements from
repeated samples across
differing microarray technologies. Quality control
analysis of microarrays. |
