Portal:Complex Systems Digital Campus/E-Laboratory on Embryome Digital Campus

e-Laboratory on Embryome Digital Campus

The scientific and societal challenge
The embryome is defined as the embryonic physiome, i.e. the multi-scale dynamics of the organism from the egg cell through its embryogenesis and organogenesis, with reference to the Physiome as define by Jim Bassingthwaighte. The concept of physiome encompasses the goals of a novel integrative and predictive biology. It seems the most appropriate to characterize the paradigm underlying the objectives of a growing interdisciplinary community gathering biologists, physicists, chemists, mathematicians and computer scientists, tackling the understanding of biological processes through their phenomenological and theoretical reconstruction from their in vivo observation. This general framework requires designing and implementing technologies, methodologies and tools for acquiring 3D+time data at all levels of living systems and extracting the relevant quantitative information further used in multi-scale analyses. This path is expected to provide in depth understanding of living systems and of their underlying processes unattained so far. We indeed are still far from understanding the basis for living systems morphogenesis, autopoiesis, ageing, homeostasis and resilience. These systemic properties cannot be understood through the mere observation and measurement of local markers and indicators. They must at some point be tackled at the level of the whole systems multi-scale dynamics. Understanding organisms’ systemic properties is a necessary condition for sustainable progress in health sciences through improved diagnosis and personalized therapies towards personalized health.

Designing and sharing new theoretical and methodological tools
Biological processes dynamics should be reconstructed from in vivo observations at the relevant spatial and temporal scales. This can be achieved by combining (i) advanced photonic microscopy, producing multimodal and multiscale large 3D+time datasets (ii) dedicated computational analyses providing new kinds of quantitative knowledge and iii) theoretical and methodological tools from dynamical systems theory, statistical physics and mechanics, machine learning theory. From the in vivo observation of biological processes to biological insights, the scientific activity goes through a chain of tasks including raw data acquisition, description, transfer, storage, processing through a pipeline of dedicated algorithms to extract the quantitative information, annotation & storage of processed data, and then data mining & analysis through interactive visualisation, categorization, quantitative comparison, prototyping. With the final goal of integrating data at the different levels of living systems’ organization, from molecules to the whole organism in animal or plant models, the biological community cannot go much further without a massive investment in a common e-environment.



This e-environment should respond to a number of expectation: i) making it easy for biologists to process their images with the best possible strategies to extract the most precise and complete quantitative information; ii) avoiding duplication of efforts; iii) making sure that the quantitative information extracted from different experimental contexts (model organism, sample preparation, imaging technology) may be compared and combined to provide biological insights; iv) making sure that the information produced is easily verifiable, reproducible, and reusable; v) readily integrating new data acquisition methods and information processing technologies and methodologies and vi) ensuring that its database framework is adequate for integrating all the relevant data including traditional “omics” data.

Embryomes-DC aims to respond to the needs of a growing interdisciplinary community. It will provide unified access to and seamless integration of the underlying networking, computing and data infrastructures and services by and for this interdisciplinary scientific e-community.

Embryomes-DC responds to the urgent need to federate individual initiatives and synergistically contribute to the achievements of a common e-infrastructure. The latter should provide a platform for “sequencing” cellular branching processes or identifying and tracking sub-cellular structures or gene expression patterns in an automated and systematic way as done for DNA, RNA and proteins. The Embryomes-DC e-environment aims at: i) attracting biologists to use the new infrastructure, rather working “on the side” with their home computer scientist; ii) attracting computer scientists to focus on contributing to this infrastructure, rather than working “on the side” with their home biologists, and iii) being a worldwide recognized concentration of knowledge for carrying out research in model organisms’ Embryomes.

Embryomes-DC ambitions to become an internationally recognized e-infrastructure concentrating expertise and services and federating currently dispersed initiatives. At the onset of the project, Embryomes-DC will offer services of customized algorithmic workflows executable on the international scientific grid-cloud and producing standardized and quality controlled annotated cell lineage trees of developing model organisms, providing the core of the Embryome. We will then progressively provide transnational access to a computing ecosystem for the automated analysis of 3D+time image datasets and knowledge integration. Autonomic computing and programming concepts allow facing the scalability challenge and the requirements of a growing number of distributed image-producing facilities and image processing and analysis centres to tackle the phenomenological reconstruction issue from molecules to whole organisms. Embryomes-DC education resources ecosystem [1], will produce and maintain an integrated map of knowledge, and thus respond to the research, training and dissemination needs of the Embryomes’ community.

of the responsible for the e-laboratory
Nadine Peyriéras | nadine.peyrieras@inaf.cnrs-gif.fr | CNRS Gif-Campus - France

list of the teams participating in the e-laboratory

 * Team CNRS Gif-Campus (Gif-sur-Yvette, France) : http://www.cnrs.fr/index.php
 * Nadine Peyriéras (Principal investigator)
 * Paul Bourgine (Principal investigator)
 * Kergosien Yannick (Professor)
 * Emmanuel Faure (Research Engineer)
 * Thierry Savy (Research Engineer)
 * Gaëlle Recher (Post Doc)
 * Barbara Rizzi (Research Engineer)
 * Julien Delile (Research Engineer)
 * Matthieu Herrmann (Research Engineer)
 * Mark Hammons (Research Engineer)
 * Monique Frain (Staff scientist)
 * Adeline Boyreau (Research Engineer)
 * Mathieu Bouyrie (PhD student)
 * Dimitri Fabrèges (PhD student)
 * Paul Villoutreix (PhD student)
 * Juan Simões (PhD student)
 * Adeline Rausch (PhD student)


 * Team Lille University (Lille - France): http://www.univ-lille1.fr/
 * Pierre Suret (Professor)


 * Team at STUBA (Bratislava - Slovakia): http://www.stuba.sk/
 * Karol Mikula (Professor)
 * Robert Cunderlik (Research Engineer)
 * Mariana Remesikova (Staff scientist)
 * Robert Spir (PhD student)
 * Smisek Michal (PhD student)


 * Team at UPM (Madrid - Spain): http://www.upm.es/
 * Andres Santos (Professor)
 * Maria J. Ledesma Carbayo (Professor)
 * David Pastor Escuredo (PhD Student)

Coordination committee

 * Nadine Peyriéras - CNRS Gif-Campus - France
 * Paul Bourgine - CNRS Gif-Campus - France
 * Pierre Suret - Lille University - France
 * Karol Mikula - STUBA University - Slovakia
 * Andres Santos - UPM - Spain

Research projects in the e-laboratory

 * Methodological and technical innovation
 * Development of a microscope “artificial scientific assistant’ based on selective plane illumination microscopy imaging, real time image processing and feedback on the imaging scheme.


 * Biological applications
 * To a large variety of morphogenetic processes in animal and vegetal model organisms.

e-laboratory Scientific Committee
(to be completed)

URL for the Website and/or Wiki of the e-laboratory

 * website of the BioEmergences platform
 * page at Wikiversity

Grid, Cloud, or other network utilities to be used

 * The computational tools are available through a unified webservice running on the BioEmergences clusters linked to the GIANT GRID. The BioEmergences clusters are used by wo Virtual Organizations (VO): i)  BioMed and ii) Complex Systems. BioEmergences clusters are exploited under the cloud computing mode, thanks to the OpenMole Operating System.


 * Access to distributed data is achieved through IRODS. The latter is standard for High Energy Physics all around the world and is currently used by the BioEmergences platform.


 * Amazon EC2 server (http://aws.amazon.com/fr/) can be used for data storage.


 * TINAsoft: http://tinasoft.eu/ for knowledge maps useful for the projects of the e-lab by studying bipartite graph between authors and theirs topics of all the relevant scientific corpus.

Data and/or Tools to be shared
At the beginning, the tools (data storage, workflow, visualisation) to be shared are the tools of the BioEmergences Platform described on the first page of its website. Then the tools to be shared will come from the whole Embryome e-community under open source development.

Enhancement of the different modules of the workflow will be done partly through partner’s addition of new modules, partly through autonomic computing and partly through international contests on gold standard data. The aim of an international contest is twofold: stimulating the imagination of scientists including young ones and assuring that the workflow at a given time is an excellent one with the required flexibility for adapting to different animal’ model and different observations protocols.

Gold standard data will be automatically shared, as well as the data used in publication (in order to allow the checking that the algorithms are well working providing correct interpretation and reconstruction of the raw data.

The Embryome-DC will recommend that the new raw data will be produced under a “delayed Creative Commons” licence, meaning that the Creative Commons licence will apply to the new data after some specified event (like, for example, their first use in a publication).

Comments
With the Embryome’s paradigm in mind, the founding partners of this e-lab [2] made foundational breakthroughs towards the multiscale reconstruction of chosen animal models’ embryogenesis [3]. They designed and implemented the prototype of a platform linked to GIANT to achieve the reconstruction of the cell lineage tree from 3D+time image data sets acquired through the in toto observation of animal model organisms (Drblíková and Mikula, 2007; Melani et al., 2007) (Krivá et al., 2009) (Zanella et al., 2009). This service is today unique and fully original worldwide. This service has a limited task force [4] and has only been dealing so far with the automated reconstruction of the embryonic cellular dynamics. Correlating with our efforts to disseminate our EC projects achievements, an as yet small number of research groups proposed in the recent years their own automated algorithmic workflows to achieve the reconstruction of cell dynamics from the observation of live model organisms. These initiatives had so far a limited impact on European Research Activity, either because strategies appeared to have very poor precision and accuracy (Keller et al., 2008) or because they were based on commercial software with limited performances (Supatto et al., 2009) or were limited to very specific types of small datasets (Fernandez et al.). Furthermore and most importantly, none of these strategies was designed to be scalable and integrated in GIANT.