DSpace@MIT

Thesis: S.M., Massachusetts Institute of Technology, Department of Biology, 2017.Cataloged from PDF version of thesis. Thesis without Abstract.Includes bibliographical references (pages 73-85).Introduction - The biological sciences have recently experienced an explosion of data unparalleled in any field, we can now survey with high accuracy, coverage and at high throughput the three key components of the cell: RNA, DNA and Amino Acids. This wealth of data has produced vast numbers of correlational studies, such as Genome Wide Association Studies (GWAS) and differential gene expression studies, but with this vast amount of data it quickly became extremely difficult, if not impossible, to digest it in a fruitful manner. Currently we can take small snippets of this whole body of data and, using specialized tools and knowledge, ask relatively simple questions. In order for the community at large to truly take advantage of the explosion of biological data now being produced we require a computing platform which is powerful, lightweight, interoperable, highly reproducible and pliable which has the capability to start from raw sequencing data, adapt, and carry the analysis all the way to designing the plasmid to transfect an organism. Although building such a platform is possible, it would only be able to guide biologists in a useful manner if they themselves are able to understand, edit, and ask questions from it. In order for this to occur we must employ a physical hierarchical model of the cell based on the underlying mechanisms within the nucleus which carry out the genetic program. If we base our model on the underlying physical reality we are able to expose the user to a subset of the hierarchies while remaining intuitive, allowing a biologist to traverse vast amounts of data and make physically informed predictions quickly.