Lets Elucidate the Cell
2009 January 24
Well this is my very first post on Cellucidate, a new web-based computational environment focused on cell signaling research introduced by Boston based Plectix BioSystems, Inc. I must accept that Cellucidate web interface is pretty cool, and stress is not just on analysis and simulation of signaling pathways, but also on sharing and annotating the whole process. Cell signaling is part of a complex system of communication, and its not possible to explore the whole signaling network alone by any single research group. A collaborative community effort is required to explore the complexity of cell signaling networks, and efforts like UCSD-Nature Signaling Gateway and Science STKE are trying to catalog the complex world of cellular signaling. Cellucidate enables scientists to answer meaningful cellular signaling “what if” questions with specificity, and simplicity using rules defined using Kappa. Cellucidate computational framework is based on Kappa language, which is inspired from key concepts in organic chemistry. In Kappa terminology signaling is described using agent (basic unit, consist a name and a number of labeled sites, a site may have a internal state), rule (rules are defined using agents and describe how agents interact) and model (described using a set of isolated rules). Once a model has been created, user can run simulations and stories. In simple words simulation provides time trace for whole model (global behavior), while story is centered around a particular rule. In Cellucidate, a new in silico signaling experiment is created as a book (see below, a normal bookshelf or collection of books),
and further chapters are added in the book, each chapter can have 4 types of pages: agent, rule, journal and simulation (see below).
Each page define only one component of respective type and normally pages of one type can be grouped under one chapter, for example in previous image for 5 agents we need to create 5 agent pages and they are grouped under Agents chapter (name of chapter can be anything not necessarily Agent). Journal pages summarize the experimental details using text and/or images. Simulation page includes contact and influence maps, along with simulation graphics.
Books can be shared in public mode, any one can copy a public book to their bookshelf and edit them. Public books in Cellucidate allows sharing of data, knowledge and models to facilitate collaboration between researchers. I hope that Cellucidate will support different standards for model import and export (currently it supports import from SBML). Cellucidate as a tool is impressive with very good look and feel, and it provides excellent workflow for collaborative cell signaling research, but at the end all modelling capabilities are centered around Kappa language. It is not clear how Kappa (rule based modelling) is better than its ODEs (Ordinary differential equations) modelling counterparts such as little b. Little b is based in Common Lisp with initial focus on systems biology, modularity and abstraction are its best features allowing widespread sharing and reuse of models, while Kappa is simple and abstract, and still in evolving stage. Kappa describe cellular signalling in terms of rules, thereby avoiding the combinatorial explosion besetting differential equations. From Plectix blog
In “sculpting mode” we inscribe behavior into a model. In “revealing mode” we read out the potential behaviors implied by a set of facts about protein-protein interactions. Ordinary differential equations (ODEs) can be useful in “revealing mode”, but only for small systems. For large systems, like in molecular signaling, modeling with ODEs easily degenerates into sculpting. Revelatory models are important for discovering and honoring the complexity of real systems. They are better – much better – tools for design (think synthetic biology) and intervention.
Let say Kappa provides an alternative to ODE based modeling, but how it will handle models with PDEs or gradients.
Updates:
- Here is the link to the paper describing Plectix’s approach to rule-based modelling of cellular signalling.
- Abstract page for rule-based modelling workshop
No related posts.
4 Responses
leave one →
I agree Kappa based approach is very much simple and best suited for biologist working with cell signaling, further cell signaling is not an isolated system, it governs basic cellular activities and coordinates cell actions, how Kappa will handle other cellular activities let say metabolic models, or it will not
@ Plectix blog – it is not clear how it is superior in this case,
Yes Anil that may be interesting, but let not make conclusions about Kappa,
Intresting stuff
Hi nice article, Kappa is simple compared to other approaches such as BlenX language which inspired from the process calculus Beta-binders