Cell-ID is an open-source cell-finding, tracking, and analysis package.
We have used it extensively to track cells through time courses of cells
containing fluorescent reporters, reporters of fluorescence translocation,
and FRET reporters (Yu et al, 2008); and we have also used the
code to accurately measure total numbers of fluorescent molecules.
The open source nature of the program has allowed us to develop and incorporate novel methods and statistics as needed. For example, we have developed a novel calculation of volume from bright field images (Gordon et al 2007), and, we have also adjusted the code to analyze the split images of our FRET experiments (Yu et al, 2008).
Cell-ID is designed to identify cells from brightfield images. This is advantageous since it avoids any bias related to the fluorescence of the reporter, such as a higher efficiency to find brighter cells or a tendency to identify brighter cells as larger. It also allows the code to identify cells even if the fluorescence distribution is highly non-uniform within the cell. However, since the code is open source, the cell-finding algorithm can be easily altered to search for cells based on fluorescence.
We have used this code primarily with yeast cells, but Cell-ID readily identifies relatively regularly shaped cells that produce dark boundaries in brightfield images. In its current implementation, it is fine-tuned to find yeast cells and is able to separate growing buds as new daughter cells while maintaining the ability to distinguish cells with large mating projections as single cells.
The code takes the pixels of each found cell and copies them over to corresponding fluorescence images, allowing for slight cell shifting of the individual cells between the bright field and fluorescence images, as well as some overall differences in image registration. It then calculates a number of single cell variables using both the fluorescence image(s) and the brightfield image (such as area, volume, fluorescence in various cellular annuli, etc). Each of the algorithms can be easily modified and other algorithms and statistics added. For time course studies, the individual labeled cells are tracked through the time course.
The output of the program is an image annotated with cell boundaries and numbers, as well as a series of tables in text format containing all the calculated data. We have been analyzing the data from such files using the open source analysis package PAW, and we have included a number of PAW scripts to aid in using PAW. Since the program output is in text format, other analysis packages (such as, R or ROOT can be easily used.
The most recent update to Cell-ID incorporates several new features and enhancements. Cell-ID 1.4 (release date October, 2008) incorporates a graphical user interface (Vcell-0.1), which provides easy testing options that help the user choose parameters to process experiments in batch mode, as well as bug fixes and improvements. This release also includes a new package, Rcell, to aid in the analysis of Cell-ID output files. Rcell contains a set of functions to load these files into R, filter out unwanted cells, display images and plot simple and compound variables.
A detailed protocol describing how to use Cell-ID to quantify cellular parameters from individual cells and track them over time will be available soon (Chernomoretz, A., Bush, A.,Yu, R.,Gordon, A., and Colman-Lerner, A.,  Using Cell-ID1.4 with R for microscope-based cytometry. Current Protocols in Molecular Biology, download).
Cell ID is made available to the public through the GNU Lesser General Public License. Please see the README file for instructions on how to install and run Cell ID.
Earlier Releases: Cell-ID1.1
Gordon A, Colman-Lerner A, Chin TE, Benjamin KR, Yu, R.C., Brent R: (2007) Single-cell quantification of molecules and rates using open-source microscope-based cytometry. Nat Methods.
Chernomoretz, A., Bush, A.,Yu, R.,Gordon, A., and Colman-Lerner, A.,  Using Cell-ID1.4 with R for microscope-based cytometry. Current Protocols In Molecular Biology.
[back to top]
Monod is an extensible knowledge management and collaboration
web application. Monod stands for MOdeler's NOtebook
and Datastore in honor of Jacob Monod, the pioneering molecular
biologist who, with Francois Jacob, abstracted the relevant
details from their experiments that enabled them to devise
the Lac operon model. Monod was designed to aid researchers
in extracting, annotating, and organizing relevant data about
biological systems and to model system behavior. The user
can input data about molecular species and the types of reactions
they undergo. The user can specify post-translational modification
states and the reactions that depend on those states. Each
of the species can be described using keywords to facilitate
searching, and species and reactions can be annotated with
user comments and linked to relevant citations. The citations
can also be annotated with the user's comments. The
Monod web application is built with open source components
including the underlying PostgresSQL 7.3 database. The user
can set access parameters to allow others access to the database.
When deployed as a shared resource Monod is a functional
data dissemination and collaboration environment. Monod was
originally designed to model an intracellular signaling system,
however it can be easily modified to include other data types
to facilitate modeling of other biological processes. Monods
extensible design also facilitates its integration into a
network of software, such as simulators and data warehouses.
Monod is available to the public through the Lesser Gnu
Soergel, D., George, B., Morgan-Linial, R., and Brent, R.
(2004) Monod, a tool to support collaborative modeling
of biological processes. [ PDF ]
1.0 is a non-spatial stochastic simulator for
cellular reaction networks. It was designed to enable a researcher
to simulate the behavior of intracellular signaling systems
over time and in response to defined perturbations. The outcomes
of the simulations can then be tested experimentally. A
unique feature of Moleculizer is that it generates species
and reactions, as they are needed. This 'on- the-fly' mode
saves time by increasing computational speed and reducing manual
input by the researcher. Moleculizer uses Monte Carlo methods
to simulate the reactions. Moleculizer models can be exported
as SBML files that can be subsequently imported into other
modeling and simulation software such as ECell.
Moleculizer 2.0 will be released in late 2006/early 2007.
Significant new features of version 2.0 include a spatial stochastic
simulator (Compartmentalizer) and replacement of loadable reaction
modules with a simplified suite of reaction generators. Moleculizer
2.0 has the capacity to generate reaction networks without
simulation. These reaction networks can then be loaded into
other simulation software.
Moleculizer is available to the public under the Lesser GNU
Lok, L and Brent, R. Automatic generation of cellular
reaction networks with Moleculizer 1.0. Nature Biotechnology,
[ back to top ]
Smoldyn. A spatial stochastic simulator for chemical reaction networks
Smoldyn is a computer program that simulates chemical reaction and diffusion systems with high accuracy, good efficiency, and spatial resolution of a nanometer or less. Smoldyn is based on and named for Smoluchowski dynamics, which is a very simple description of chemical reaction dynamics. In Smoldyn, every molecule of interest is represented as an individual point, while those that are not of interest (water, non-reactive molecules, etc.) are not represented. Simulated space is continuous, meaning that space is neither a lattice, nor binned into voxels. The use of continuous space provides good spatial resolution and avoids artifacts that can arise from lattice geometries. Space can be 1, 2, or 3-dimensional.
To address the needs of many systems biologists, recent additions have focused on membrane-associated components and processes. These include transiently membrane-localized proteins (e.g. Ste5), trans-membrane receptors (e.g. Ste2), ion channels and pumps, and semi-permeable membranes. Each membrane can be an arbitrarily complex triangulated mesh surface, or as simple as a single sphere.
Smoldyn is developed and maintained by Steven Andrews. It runs natively on MacOS 10.4.8 (Intel) and can be configured to run Linux and Windows OS (requires OpenGL). All of the files necessary to install and use Smoldyn (e.g. binaries, source code, makefiles, utilities, and documentation for users and programmers) can be downloaded from www.smoldyn.org.
Andrews, Steven S. and Bray, Dennis. (2004) Stochastic simulation of chemical reactions with spatial resolution and single molecule detail. Phys. Biol. 1:137-151.[PubMed]
Andrews, Steven S. (2005). Serial rebinding of ligands to clustered receptors as exemplified by bacterial chemotaxis. Phys. Biol. 2:111-122. [PubMed]
Lipkow, Karen, Andrews, Steven S., and Bray, Dennis. (2005). Simulated diffusion of phosphorylated CheY through the cytoplasm of Escherichia coli. J. Bact., 187:45-53. [PubMed]
Lipkow, Karen. (2006). Changing cellular location of CheZ predicted by molecular simulations. PLoS Comp. Biol. 2:0301-0310.[PubMed]
Andrews, Steven S. and Adam P. Arkin. (2006) Simulating cell biology Curr. Biol. 16:R523-R527.[PubMed]
HillSim. Dose Respomse optimization for signaling systems.
HillSim is a computer program for analyzing and optimizing signaling system models. An input, nodes, and arrows comprise signaling system models: the input is typically some extra-cellular chemical, such as the alpha-factor pheromone for the yeast pheromone response system; nodes typically represent signaling proteins, and have activities that can vary from fully inactive to fully active; and arrows create causal links between the nodes so that signaling proteins can activate or inactivate each other. Arbitrarily complex network topologies can be modeled, including those with feedbacks, feedforwards, or arrows that multiply the activities of multiple nodes.
While HillSim can simulate signaling system dynamics, it is best for analyzing their steady-state properties. HillSim calculates steady-state dose-response behaviors for system nodes, fits them to Hill functions, and compares these Hill function fits to “target” dose-response behaviors that are provided by the user. HillSim can also optimize the signaling system parameters so that the model dose-responses match the target dose-responses as closely as possible.
HillSim is developed and maintained by Steve Andrews. It runs on Mac OS X, Linux, and Windows, and is available to the public under the GNU Lesser General Public License. All of the files necessary to install and use HillSim (e.g. binaries, source code, makefiles, and documentation for users and programmers) can be downloaded here.
Andrews, Steven S., Yu, Richard. C., and Brent, Roger. (2008). Cell signaling dose response alignment may arise from cooperativity or pull-up/pull-down mechanisms. Molecular Systems Biology, submitted.
[ back to top ]