Metachem

Obiettivo

A. GENERAL BACKGROUND

A1: Why a COST Action on this topic?

Modern computational Chemistry is moving towards more advanced computing tools because of the increasing need for realistic simulations of complex systems relevant to the modelling of several modern technologies and environmental phenomena. Such realistic simulations usually need to include, although not necessarily in a completely rigorous manner, a detailed (possibly a priori) description of the relevant chemical substances and processes. As a result, the computational procedures of modern chemical simulations are quite complex, and the related computer codes must be run on extremely powerful machines. This means also that most often they cannot be implemented on the computer of the user but need to reside on a remote site and to be used via the Internet. Moreover, this type of applications frequently requires the assembly of expertise from various areas as well as various hardware and software components.

As far as co-ordinating the necessary expertise is concerned, COST has a long standing tradition and has already proved to be a viable and successful approach to scientific and technical co-operation in Europe especially in the field of Chemistry.

To limit ourselves to advance computational Chemistry one can recall that through D3 (1992-97) and D9 (1997-2002) COST based collaboration among European laboratories has been very successful in the a priori modelling of electronic structure and molecular dynamics. This includes significant advances in such areas as the treatment of increasingly heavier particles and larger molecules, the inclusion of all coupled degrees of freedom and multi-potential energy surfaces, the simulation of gas and condensed phase processes.

With respect to linking various pieces of hardware and software, this becomes of particular importance for scientific, economic and managing reasons when computational applications move to a higher level of complexity and can no longer be hosted in a single location. The proposed solution is the establishment of a METALABORATORY, i.e. a cluster of geographically distributed resources. Metalaboratories would facilitate the pooling of hardware, software and even "brainware" by establishing collaborations between scientists having complementary expertise and access to complementary but compatible resources. For a large scale metalaboratory, one can envision parts of the computational application running on one or more computers (individually owned by the participating laboratories) linked in a centrally managed metacomputer system.

The proposed COST action D 23 is intended at fostering the amalgamation of European Laboratories working on common complex chemical applications into Metalaboratories assembled around a Metacomputer system. The transdisciplinary collaboration between chemists and computer scientists is aimed also at reducing the gap in the development of computing tools for metacomputers especially for the benefit of user of Chemistry.

A2: Status of the research in this field

A2.1 The evolution of complex computational chemical applications

At present, several software packages devoted to specific a priori calculations of properties of chemical systems are available. Significant progress has been made towards the calculation of electronic energies and potential energy surfaces for large molecules or ensembles of atoms and molecules in gas phase, in solution and in solid state to be combined with quantum and/or classical dynamics treatments as well as with statistical approaches to reproduce quantities measured in experimental Chemistry.

These experiments are concerned with magnetic resonance, photophysics and photochemistry, molecular beams, reaction rates and cross sections, magnetic, optical and electric measurements. Other computational applications are concerned with synthesis planning, drug design, harmacophore discovery, biological activity evaluation. Some of the relevant programs are [1,2] AMBER, CHARMM, CHEMICS, COBRA, COLUMBUS, CONCORD, CORINA, DISCOVER, ECEPP, EROS, GAMESS, GAUSSIAN, GROMOS, HONDO, MMFF, SESAMI, WIZARD, WODCA.

All these software products are increasingly associated with computer science products like hypermedia applications, artificial intelligence tools, web and networking facilities.

The present trend, in fact, consists of moving towards realistic computational applications in which all the components of a real system are included into the model. This implies the involvement of a wide spectrum of competence and the assemblage of truly complex computer codes based upon existing academic commercial codes as the design and the development of new software components as well as the navigation in the web and the construction of highly interactive interfaces.

A2.2 The evolution of concurrent computing

Most of the existing computational Chemistry applications already exploit the innovative features of parallel computers. Parallel architectures are able, in fact, to offer the extra power needed for running most of the existing computational applications. However, single machine architectures have intrinsic technical limits and frequently show a less favourable price/performance ratio.

Heterogeneous distributed computing (metacomputing) is a fast growing area in high performance computing but it is also the most cost effective approach to intensive computations thanks to the use of the best suited hardware platform and to the use of a large variety of networks. A metacomputer is a computing platform made of a geographically distributed cluster of heterogeneous computers connected on a network (LAN, MAN or WAN) through a software that co-ordinates them as a single virtual parallel machine [3]. These systems can be made of workstations, compute servers, massively parallel computers (MPPs), specialised computing equipment, file servers, mass storage robots, data bases, networked experimental apparatuses, etc. They are addressed to computer applications whose complexity makes the use of single computer architectures inefficient and allow the full utilisation of unused machines cycles. Centrally managed clusters of heterogeneous computers feature also the advantage of encouraging design, development and concerted use of proprietary, commercial and shared software.

They also favour the use of chemical computational applications by third parties.

Software tools used for metacomputing can act either at low level (like Condor[3] that manages the distribution of computational applications over the whole metacomputing system) or at a higher level (like Globus[4] that manages the various resources through the web). Other tools such as message passing interfaces, metacomputer schedulers, job flow management and application steering procedures combine these features with other additional useful characteristics.

References

[1] Activity report 1995-96 of COST Chemistry Actions, Luxenbourg, 1998 ISBN 92-828-2603-1; Activity report 1997-98 of COST Chemistry Actions, Luxenbourg, 1999, ISBN 92-828-7416-8.
[2] Encyclopedia of Computational Chemistry, P. Von Rague' Slayer, N.L. Allinger, T. Clark, G. Gasteiger, P.A. Kollman, H.F. Schaefer III Eds. (John Wiley, Chichester, 1999).
[3] I. Foster, C. Kesselman Eds., The Grid: Blueprint for a New Computing Infrastructure, Morgan Kaufmann Publ., San Francisco, 1999,
[4] L. Smarr, C.E. Catlett, Metacomputing, Communications of the ACM, 35(6), 45 (1992).

A3: Relationship with other European Programmes and Industrial Initiatives in Europe

The present action is in many aspects a gemmation of the current action D9. It does not cover all aspects of D9 since it focuses on computationally intensive applications which need, because of their complexity, the pooling of various competence and the availability of extra computing power. As far as competence is concerned there are in Europe several initiatives (Conferences, projects, Schools, etc.) and computational Chemistry laboratories whose activities are focused on the assemblage of complex computational procedures to estimate the characteristics of electronic structures, collision dynamics, reaction rates, photoionization processes, gas and condensed phase kinetics. In Europe there is also competence in managing metacomputer systems and developing hypermedia computer tools for their effective use. This suggests a strong interaction with COST initiatives in the field of electronics and informatics.

Within the perspective of moving to more complex computational applications more and more appreciative fields need to be considered in the present initiative. This is of particular interest when considering all the aspects of University- Industry relationships and considering all the aspects of the synergy between various European initiatives. At this stage, the project CAPRI (Competitive Advantage through Process Information Technology) developped in the industrial collaborative programme SUSTECH (Sustainable Technology in Chemistry) promoted by the CEFIC (Council of European Federations of Chemical Industries) would be particularly attractive to organised co-ordinated research efforts.

A particular emphasis should be put into links with COST initiatives in computing and 5th framework programme like the one for the user friendly computing. This is of particular interest for the proposed action since the institution of metalaboratories and the remote use of metacomputer machines as a fundamental step the usage of web pages.

B. OBJECTIVES OF THE ACTION AND SCIENTIFIC CONTENT

B1: Main objective

The main objective of the Action is the establishment of more effective collaboration mechanisms among European computational Chemistry laboratories through the constitution of METALABORATORIES (clusters of geographically distributed laboratories having complementary expertise on specific chemical applications working by sharing computing resources in a meta-computing system) to develop complex computational applications in Chemistry and related computing tools.

This will be achieved by establishing a novel co-operation model aimed at developing the a priori modelling of complex chemical systems and encompassing the necessary scientific competence and/or computing means from different laboratories.

It would be possible to assemble chemical simulations that cannot be properly implemented in a single laboratory due to the critical size required in terms of expertise, manpower, infrastructures and organisation necessary to succeed in managing the different aspects of the related computational procedures. This also prevents possible developers and potential users of these simulations from allocating financial resources to related R&D projects. The establishment of Metalaboratories, instead, provides a protected environment for developing new computational solutions without affording individually the full cost and without needing to acquire the whole amount of know how. It also provides a means for final users (like private companies and software houses) for collaborating with creative academic research and get access to its products.

In particular, some scientific goals of the project could be:

1. Establish in Europe Metalaboratories as an advanced experience of co-operation to design and develop computationally intensive chemical projects.

As example the modelling of molecular processing on a priori basis like describing what is happening in a solute-solvent system need to develop new algorithm for the computers and to create new chemical approaches of molecular dynamics.

2. Enhance research development in computational modelling by going beyond the limitations of single machine approaches i.e. by adopting a metacomputing approach sharing software and computing platforms.

For developping new algorithms for chemical applications, using the possible calculation of hypersurfaces would need large repetitive calculations which normally are performed on parallel computers. But the alternative approach with metacomputing improves the effficiency of computing the hypersurfaces by performing a variety of tasks, such as large scale ab initio calculations combined with the concurrent testing by classical simulation methods and optimisation of the hypothetical model potentials, at the same time on different computing platforms.

3. Construct web interfaces and computer tools for a friendly use of shared resources;

The chemist has to use the machine in the cluster which is the best available for his/her problem. The software to assist him in this choice has to be " chemist intelligent " i.e. it has to be acting as if it thinks as a " Chemist ".

4. Set a working scheme for extending cost effective distributed computing experiences in Europe to active research communities (like the Chemistry community) able to lower the investment threshold for advanced computational applications;

Already in Europe several Large Scale Computing Facilities are devoted significant part of the time of use to Chemistry :
- Manno Computing Centre in Switzerland ;
- Cinecca in Bologna, Italy ;
- IPCC at Edimburgh, United Kingdom ;
- Desbury Laboratory, United Kingdom ;
- CEPBA-CESCA in Barcelona, Spain ;
- IDRIS in France.

5. Bring closer to the end-user and to the production world computational academic research.

The promotion of high level computational Chemistry ( especially ab initio and DFT for interface in virtual reality problem) will be a particular subject for teaching in area such as :
- processing of nuclear waste through virtual capacity ;
- simulating waste production and possible treatment of chemical production.

B2: Metacomputing for Chemistry

There are different types of problems opened by the Metacomputer based organisation of a Metalaboratory. Other problems originate from the different aspects involved in the management of a Metacomputer when this is applied to large chemical applications that may interest a potentially wide area of subjects.

B2.1 Metacomputer management

Metacomputer management aspects are concerned with the design and implementation of tools necessary for the management of the clustered machines, with the improvement of networking capabilities, with the harmonisation of the software, with testing and alignment of its different versions, with the distribution of the running versions, with the sharing of proprietary and commercial software and with the implementation and use of the application on the web.

B2.2 Applicative aspects

There are large families of chemical computational applications that can profit from the development of Metalaboratories for their full implementation. Some of the fields, without pretending to be exhaustive, are green Sustainable Chemistry, life Chemistry, medicinal Chemistry, material Chemistry, chemical education, etc.

Examples of possible Metalaboratories that could be developed within the proposed COST action in the above mentioned fields are:

a. " Software package maintenance and development "

Many of the computational packages mentioned in section A originated from academic research. The need of assembling computer codes able to tackle thematic research areas has been recognised by the Chemistry community since the very beginning of the computer era. Presently there are several packages that need to be maintained and continuously updated. This effort would be more easy to sustain in a Metalaboratory. A metalaboratory would make easier to manage also the use from third parties and the construction of specific web interfaces.

b. " Collaborative experiments management "

Modern experiments more and more require the joint effort of laboratories running remote installations and laboratories developing complex management and interpretation software. The concertation of these efforts as well as the development of related software in an ideal test bed for developing Metalaboratories

c. " Field development "

Advanced computational studies require the development of new approaches and algorithms for giving a more realist description of chemical substances and processes. The final target of most of these studies is the construction of a virtual reality based on molecular frames (molecular virtual reality) for which the complexity of chemical laws need to be properly dealt by the concerted action of several laboratories. Such a co-operative effort is a suitable platform for a Metalaboratory.

d. " Application development "

The assemblage of academic and commercially goals at innovative educational, industrial and social applications may be properly carried out by Metalaboratories of mixed academic nature. In these cases non-chemical competence may be needed in addition to chemical ones. In these particular cases additional attention should be paid to the characteristics of the interfaces and of the network to bring in security.

B3 Scientific programme and prerequisites

The scientific programme will depend on the projects submitted by individual research teams. The working group projects will be selected according to the objectives outlined above and to the establishment of a Metalaboratory through the assignment of a machine of the proper level to the metacomputer system. Each Metalaboratory will include, at least, a particularly skilled laboratory or a computer centre committed to manage the metacomputer.

C. ORGANISATION, MANAGEMENT AND RESPONSIBILITIES

C1: Organisation and Management

Research projects fitting topics mentioned above will be submitted to the Management committee for approval. Approval will be subject to the evaluation of their relevance for the establishment of Metalaboratories for computational chemical applications of interest for the European member countries.

C2: Responsibilities

The management committee has responsibilities for:
1. Drawing up the inventory during the first year, organisation of workshops and start of the activity; existing contacts will be used which should greatly facilitate this task.
2. Co-ordination of the joint activities with other COST Actions, CEFIC Sustech Clusters; joint meetings are likely to result from this activity;
3. Explore the possibilities for wider participation and exchange of information with EU-specific programmes, ESF, EUREKA, etc;
4. Planning the intermediate report, the final report and the concluding symposium.

Progress reports will also be provided by each respective participant in the projects in their own countries within the framework of existing programmes.

C3: Evaluation of Progress

The progress of the programme will be monitored by means of brief annual reports from each of the participating scientists. These will describe the results of research obtained through concerted action. The Management Committee will prepare a milestone report after 2 years of joint activities. The report will be presented to the COST Technical Committee for Chemistry for their review and to the COST Senior Officials Committee for information. A final report will be published to inform
Non-participating scientists and research workers interested in the results about the scientific achievements of the Action. It is expected that some review by participants, which describe the progress, made and state of the field, will be published in International Journals. To conclude the COST Action, a symposium will be held after 5 years. It will be accessible to other scientists and potential users of industry.

D. TIME-TABLE

The programme will cover five years and consist of the following stages:

Stage 1:After the first meeting of the Management Committee a detailed inventory of on-going research and existing plans of the participating groups to begin joint projects will be made. This will result in a discussion document to allow further planning.
Stage 2:It will be evident which projects are closely related and would benefit from joint activities. Researchers (and co-workers) will set-up (and continue) joint collaborative projects and exchange their recent research results. It may be appropriate to explore wider collaboration with other European countries during this stage.
Stage 3:An intermediate progress report will be prepared after two years for review by the COST Technical Committee for Chemistry and for information to the COST Senior Officials Committee.
Stage 4:This final phase will begin after four years and will involve the evaluation of the results obtained. It may include the organisation of a symposium for all the participants and co-workers. The final report will be submitted to the COST Technical Committee for Chemistry for scientific assessment and after to the COST Senior Officials Committee.

In summary the total timetable can be represented as follows:

- Start 1st year
- Formation of projects 1st and 2nd year
- Workshop of group leaders end of 1st year and each year later on
- Overview available; start meetings; continue meetings on subtopics end of 2nd year
- Start exploration of wider participation 3rd and 4th year
- Intermediate Progress Report available for Technical Committee and CSO end 2nd year
- Start evaluation of results continuously each year after 1st year at the yearly workshop
- Concluding Symposium end of 5th year

E. DURATION OF THE ACTION

The Action will last for five years

F. ECONOMIC DIMENSION OF THE ACTION

It is to be expected that several teams of the presently running COST D9 Action as well as some of the groups active in software production in other actions will apply for admission to the new COST Action. It is estimated that a total of 20 laboratories will eventually be involved in the Action. Based on experience in the COST D9 programme, it is estimated that the economic dimension of the Action (initial estimate of total costs = personnel + operational + running + Commission costs) will be 50 M _.
The total human effort in the Action "Metachem: Metalaboratories for complex computational applications in Chemistry" described above, amounts to 400 man-years (80 researchers during 5 years), being equivalent to 40 M _.

F1. Personnel costs (research and administration) will be approximately 40 M _ (covered by participant groups).

F2. Operational and running costs
The estimate of the total operational and running costs including costs of instruments and materials is 10 M _ (covered by participant groups). The calculation costs are not included in this estimation. All of them would be covered by the different participants directly.

F3. Co-ordination costs
The costs for co-ordination to be covered by the COST budget are estimated to be 0.5 M _ (100 000 _ per year) equivalent to 1% of the overall cost.

G. DISSEMINATION OF SCIENTIFIC RESULTS

All publications arising from research carried out under COST Action D23 will credit COST support and the Management Committee will encourage and promote all co-authored papers. Results of research carried out by the working groups under COST Action D23 will be submitted to international scientific journals and reviewers.

Joint meetings among different working groups in COST Action D23 and with working groups from other COST Actions, will be organised so as to best promote interdisciplinary communication.

The Management Committee (MC), in conjunction with the working groups (WG) of the Action will meet every year with the main aim of presenting results to the MC as a whole and, where possible, the MC will invite potential users and interested parties to this meeting.

The Management Committee will, during the first year of the Action, also set up a work-plan for interdisciplinary events for the dissemination of results of the Action COST D23.

Programma(i)

IC-COST - European cooperation in the field of scientific and technical research (COST), 1971-

Argomento(i)

13 - Chemistry

Invito a presentare proposte

Data not available

Meccanismo di finanziamento

Data not available

Coordinatore

N/A

Contributo UE

Nessun dato

Indirizzo

Costo totale