Skip to content

Quick Start Guide

Jump to bottom
michaelkyu edited this page Mar 6, 2018 · 39 revisions

DCell is a neural network model of budding yeast, a basic eukaryotic cell. The model structure corresponds exactly to a hierarchy of 2,526 cellular subsystems. Given this neural network structure, DCell has been trained to translate genotype to phenotype. Given a genotype consisting of a set of deleted genes in S. cerevisiae, DCell will predict the cell growth phenotype and explain how this phenotype arises from changes in the states of cellular subsystems. For example in the screenshot, changes to the states of mitochondrial respiratory chain and its sub-components explain the cell growth phenotype for the genotype cyt1Δcox7Δ.

The following steps will simulate the cell growth phenotype for the genotype cyt1Δcox7Δ, producing the above screenshot.

  1. Start DCell application. Go to http://d-cell.ucsd.edu in your browser. Click on the "Start App" button.

  2. Search for genes. In the Genotype Builder search box, type in "CYT1 COX7" and click the search symbol. The search result will show the genes CYT1 and COX7.

    In general, you may search for genes by their symbol (e.g. CYT1), the systematic name in the Saccharomyces Genome Database (e.g. YOR065W), or by keywords matching the gene's description (e.g. Cytochrome c1). You may also search for the description of a particular subsystem (GO term), such as "mitochondrial respiratory chain", to get all genes that function in that term.

  3. Define genotype by selecting genes to delete. In the search results, select the check boxes for CYT1 and COX7 to indicate that these genes are deleted in the genotype to be simulated.

  4. Simulate genotype-phenotype translation. Click on the Simulate button to simulate a genotype-phenotype translation.

    Once simulation is done, an interactive network is displayed. It shows routes from the deleted genes through a hierarchy of cellular subsystems, including mitochondrial respiratory chain, ending in the root node of the hierarchy representing cell growth. The color of each node represents the change in state of a cellular subsystem from its state in a wildtype genotype. The color of the root node represents the resulting change in growth rate

The DCell web application currently supports the simulation of cell growth phenotype using a hierarchy of cellular subsystems defined by the Gene Ontology. The web application does not yet support other simulations, i.e. the use of a data-driven hierarchy called CliXO and the simulation of genetic interaction phenotype, which are described in the DCell paper (Ma et al. 2018).

News

  • 3/5/2018 - Quick Start Guide updated for DCell version 1.4.0.
  • 8/28/2017 - Quick Start Guide updated for DCell version 1.2.0.

License

DCell is open-source software distributed under an MIT License

Questions?

Please send your questions to kono@ucsd.edu.

© 2017 UC, San Diego Trey Ideker Lab

Developed and maintained by Keiichiro Ono.

Clone this wiki locally