Christine M. Povinelli, Ph.D.
Computational Detection of Chromosome Structure and Function


Christine M. Povinelli


Contact Chris Povinelli:
cmpov@swbell.net
Skype: christine.povinelli


Ph.D. 1987
Georgia Institute of Technology
Molecular Biology and Genetics



OVERVIEW

For management and protection, the large amounts of DNA in animal cells are highly folded into compact storage structures called chromosomes.

My research asks whether information involved in gene control is precisely organized in the chromosomes of eukaryotes, whether this organization describes how DNA is folded in chromosomes, and how defects in this organization may contribute to human genetic disease.

For example, in eukaryotic 30nm chromatin fibers, the potential organization of multiple regulatory elements for two different genes might be simplistically visualized as shown in this illustration :



In order for a gene to be expressed, the DNA containing the gene, and the gene's regulatory regions, must be unfolded from highly condensed chromosome structure. In addition, several gene regulatory regions must often locate and physically interact with each other through bound gene regulatory proteins. These multiple regulatory regions, involved in the control of one gene, may be highly separated along the linear DNA.


Since the unfolding of chromatin structure and the interaction of distant gene control regions are both hallmarks of the early stages of gene activation, it is possible that gene control regions are ‘pre-organized’ in relatively close proximity in folded 30nm fibers.




As these structures begin to open and unfold, gene regulatory proteins could, potentially, bind and interact with all of the relatively localized regulatory regions, before the DNA is in a more unfolded, linear, form. Detection of periodically  distributed information, with a period length equivalent to one turn of a 30nm fiber, could therefore, describe both structural (i.e. circumference) and functional (gene regulatory) aspects of the 30nm fiber.







Click Here
Click Here
Click Here
Power Point Presentation
Contains animations
Run as a slide show

Structure and Function of 30nm Fibers Deduced From the Linear Distribution of Genetic Information.

Part One: The Logical Framework

A brief and simplified overview of the research approaches.
Short video made using MATLAB:
A Graphical Illustration of Periodicity Detection

This video shows how information (red dots) that is linearly spaced, with a period of 1200BP, is distributed when it is coiled at various circumferences. The relative locations of the information at coil circumferences between 1325 and 575 BP are shown.

The information is vertically aligned only at circumferences of 1200 BP (every rung) and 600 BP (every other rung). The occurrence of vertically aligned information is most densely aligned at the true period of 1200 BP.

Note that this type of analysis can be used both to detect the degree to which linear information is periodically distributed and, separately, to model the relative locations of information in 3D solenoidal data sets.
Short video made using MATLAB:
A Graphical Illustration of Periodicity Detection in a Split Signal with Some Noise

This short video is the same as the one to the left but with simple 1200 BP linear periodicity imprecisely split by approximately 200 BP.

The underlying period of 1200 BP is still detected by noting the circumference at which the red dots are most tightly, vertically, aligned.





SIGNIFICANCE
Health:  Demonstration that gene control sequences are precisely located in three dimensions in chromosomes would be a strong indication that this organization is important for some process, the most likely of these is for proper gene expression. This research is developing unique computational tools for classifying the types of aberrations in three-dimensional organization that can lead to genetic disease. The insights derived from these studies can then be used to design approaches for their correction.

Bioinformatics:  Three-dimensional analysis of human genetic information represents an entirely new field of study that can be referred to as‘Chromosomics’- the study of chromosome structure and function based on the distribution of genomic sequence data.

This is a profound break from the current linear methods used to display and computationally analyze genomic data. It is analogous to how protein biologists only gain a complete understanding of protein structure and function, when they determine how the linear string of amino acids is folded in three dimensions. Development of computational approaches for studying the structure of human chromosomes is necessary since chromosome complexity and fragility has made them difficult to define experimentally.

Computational and Applied Mathematics:  This data set, and the methods developed during the course of this research for its characterization, represent a completely new area of applied computational research. Prior to this research there were no methods available for examining these types of questions.