Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
G Reports from Breakout Session Groups A key component of the Workshop on Information and Communications was the set of four breakout sessions that enabled individual input by workshop participants on the four themes of the workshop: discovery, interfaces, challenges, and infrastructure. Each breakout session was guided by a facilitator and by the expertise of the individuals as well as the content of the plenary sessions (Table G-1~. Participants were assigned one of three groups on a random basis, although individuals from the same institution were assigned to different breakout groups. Each breakout group (color-coded as red, yellow, and green) was asked to ad- dress the same set of questions and provide answers to the questions, including prioritization of the voting to determine which topics the group concluded were most important. After every breakout session, each group reported the results of its discussion in plenary session. The committee has attempted in this report to integrate the information gath- ered in the breakout sessions and to use it as the basis for the findings contained herein. When the breakout groups reported votes for prioritizing their conclu- sions, the votes are shown parenthetically in this section. 185
186 TABLE G-1 Organization of Breakout Sessions. APPENDIX G Breakout Session Group Facilitator Session Chair Rapporteur Discovery Red D. Raber J. Ottino P. Jurs Yellow P. Cummings J. Tully E. Vera J. Hempel Green J. Jackiw K. Lipkowitz C. Breneman Interfaces Red P. Cummings J. Tully G. McRea J. Hempel Yellow J. Jackiw K. Lipkowitz R. Friesner Green D. Raber J. Ottino P. Westmoreland Challenges Red J. Jackiw K. Lipkowitz G. McRea Yellow D. Raber J. Ottino L. Rahn G. Martyna Green P. Cummings J. Tully P. Westmoreland J. Hempel Infrastructure Yellow-Red D. Raber J. Tully M. Tuckerbag (with G. McRea) Green-Red J. Jackiw K. Lipkowitz P. Gund J. Ottino C. Breneman DISCOVERY What major discoveries or advances related to information and communica- tions have been made in the chemical sciences during the last several decades? Red Group Report Prioritized List: Analytical instrumentation and data visualization (11) Computational materials science (10) Process design and optimization (6) Drug design and bioinformatics (5) Environmental chemistry and modeling (5) Materials applications (4) Advances in quantum mechanics (3)
APPENDIX G Yellow Group Report Molecular Simulation of Chemical Systems 187 Materials, polymers, advanced materials, crystals and nanostructures, simu- lation techniques (DFT, Monte Carlo and MD) Microfabrication by Chemical Methods (e.g., lithography) Synthesis, design, and processing of new materials Computer-Aided Drug Design QSAR, molecular modeling Numerical Simulation of Complex Chemical Processes Involving Reaction Transport Ozone layer, combustion, atmospheric chemistry, chemical vapor deposition Process Simulation, Optimization, and Control Supply-chain optimization, plant design, and remote control of plants Green Group Report Electronic Structure Theory (9) Common language of electronic structure theory, validating density func- tional theory, standardization of Gaussian basis sets, semiempirical methods, B3LYP functional (Becke's three-parameter hybrid functional using the LYP correlation functional), well-defined procedures for QM algorithms Research Simulation Software for All Science (8) Gaussian software, archiving chemical information, software development for the community, process control and process design software Computer-Aided Molecular Design (7) Combinatorial chemistry for drug study
188 Chemical Kinetics and Dynamics (5) APPENDIX G (coupling electronic structure theory with dynamics (direct dynamics), kinet- ics from first-principles rate constants, stochastic simulation theory Potential Functions and Sampling Methods (4) Monte Carlo and molecular dynamics for statistical thermodynamics, poten- tial energy functions and applications, discovery of applicability of potential en- ergy functions, reactive empirical bond order potentials Data-Driven Screening (2) Application of statistical learning theory and QSPR, data mining and combi- natorial high-throughput methods, virtual high-throughput screening Synthesis of Materials for Computing (1) Photolithography polymers, metallization, optical fibers INTERFACES What are the major computing-related discoveries and challenges at the in- terfaces between chemistry-chemical engineering and other disciplines, includ- ing biology, environmental science, information science, materials science, and physics? Red Group Report Targeted Design How do we go backward from the desired specifications and functions to identify the molecules and processes? A Matrix View Is Needed Rows (discovery opportunities): biosytems modeling, proteins, drug design, materials design, environmental modeling (life-cycle analysis), molecular elec- tronics, nanosciences Columns (challenges to achieving discoveries): combinatorial search algo- rithms, data assimilation, multiscale modeling, structure-function and structure- activity relationships, large-scale computation, how to create interdisciplinary teams as well as support the critical need for disciplinary research
APPENDIX G 189 Results: the matrix is dense (every cell is a yes); people issues are critical; opportunities exist for both methodology development and application; there are many rows and columns, so the challenges are prioritization and maximizing the intersections; we need to show relevance, examples and benefits of what we are trying to do. Yellow Group Report Molecular Design (8) Modeling at an atomic scale, small molecules vs. larger structures (one level of multiscale modeling); examples include design of drugs, nanostructures, cata- lysts, etc. Materials Design (18) More complex structures, additional length scales, property prediction (e.g., microelectronics devices, microfabrication, polymers, drug delivery systems, chemical sensors) Multiscale Modeling (17) Incorporates all length scales, important in all disciplines; examples include modeling of a cell, process engineering, climate modeling, brain modeling, envi- ronment modeling, combustion Data Analysis and Visualization (6) Analysis of high throughput data, seamlessness of large datasets, modeling of high-throughput datasets, information extraction, component architectures for visualization and computing New Methods Development (6) Computer Science and Applied Math Green Group Report Human-computer interface, expert systems, software engineering, high-per- formance computing, databases, visualization
190 Environmental Science APPENDIX G Fate of chemicals, computational toxicology, climate change, speciation Biology Protein structure-function; protein sequence-structure; networks (kinetic, regulatory, signaling); bioinformatics; dynamics in solvent environment Physics Chemical dynamics, scattering theory, molecular astrophysics, high-tempera- ture superconductors Materials Science and Engineering Polymer property prediction, large-systems modeling (mesoscale simula- tion), nanostructures, corrosion Earth Science Water chemistry; binding of chemical species to ill-defined substrates; high- temperature, high-pressure chemistry; atmospheric chemistry Crosscutting Issues The issues are pervasive: every matrix entry is a yes. Matrix columns (computing-related problem areas): software and interfaces (expert systems, development and maintenance and incentives); systems biology (chemistry, physics, and processes complexity); molecular dynamics with quan- titative reactive potentials; chemical and physical environment of calculations (e.g., solvation); simulation-driven experiments; special-purpose computing hard- ware; closing the loop for computational chemistry with analytical instruments and methods; database access, assessment, and traceability Matrix rows (disciplines that interact with chemistry-chemical engineering): computational science and mathematics; environmental science; biology; phys- ics; materials science and engineering; earth sciences CHALLENGES What are the information and communications grand challenges in the chemi cat sciences and engineering?
APPENDIX G Key Message Reducing the time to solution Some General Points 191 Red Group Report There are many possible information and communications challenges; soci- ety: no chemicals in this product; education: communicating excitement; techni- cal: information gathering and exchange; new disciplines can emerge or have emerged from taking a systematic approach to knowledge storage and representa- tion (e.g., genomics); if we are to be successful in our recommendations we need to address the challenges, opportunities and benefits (particularly with examples). A Matrix View Rows: grand challenge problems; columns: information and knowledge- based methodologies Discovery Opportunities (Rows) Simulating complex systems, cell, combustion, polyelectrolytes and com- plex fluids, atmospheres, hydrogeology, catalysts under industrial conditions, drug design, protein-DNA interactions, protein-protein, RNA folding, protein folding, metabolic networks, conformational sampling Knowledge Issues Knowledge database and intelligent agents; assimilating sensor-measure- ment-information data, data exchange standards (XML, etch; representation of uncertainties in databases, building and supporting physical property archives (NIST, Webbook, etch; collaborative environments Methodology Issues Model-based experimental design, long time scale dynamics, virtual mea- surements, force field validation and robustness, quantum dynamics, dispersion, excited states Systems Approaches to Complex Problems Intelligent and distributed databases (the "chemistry Google"), next-genera-
92 APPENDIX G tion "laboratory notebooks" for models, databases, programs, and experimental data, next-generation databases, communication, sampling, and interaction across scales; this is not a new issue. YELLOW GROUP REPORT Long-Term Challenges Comprehensive data and computational problem-solving environment (12~: seamless chemical information, computer human interface Simultaneous product and process development (12~: for example, drug dis- covery, cradle-to-grave chemical production Design of new molecule given required functionality (11~: design lowest- energy consumption path, proof of concept for computational drug design Virtual cell (4) Design biocompatible materials (3~: artificial tissue and bone, artificial or- gan, posttranslational protein modification Short Term Challenges Design-controlled directed self-assembly (10) Self-assembly of protein in membrane (5) Chemical plant on a chip (4) Predict crystal structures (small organics) (4) Computational electrochemistry (4) Accurate free-energy calculation of medium size molecules (3) Green Group Report Molecularly Based Modeling Electronic structure methods for 104+ atoms, binding energies in arbitrarily complex materials, molecular processes in complex environments, computational tools to study chemical reactions (first principles, limits of statistical reaction rate theories), simulators with quantum dynamics for quantum degrees of freedom Systems Issues Integration of applied math and computer science to obtain more powerful software tools for the chemical sciences and engineering, modeling the origin and development of life, alternative chemical bases for life (other worlds), computa- tional disease modeling, design of big-inspired materials, global climate control
APPENDIX G 193 (the earth as a chemical factory similar to Groves' chip as a chemical factory), modeling of atmospheric chemistry Measurements Incorporation of computational chemistry methods into instruments, virtual measurement, enabling ubiquitous sensing, in vivo biosensors Education and Usage Issues Translate molecularly based modeling into education, expert systems for method choices (precision in context of time and cost) INFRASTRUCTURE What are the issues at the intersection of computing and the chemical sol . ences for which there are structural challenges and opportunities in teaching, research, equipment, codes and software, facilities, and personnel? Yellow-Red Group Report Topics To Be Addressed Teaching, research, equipment, codes and software, facilities, personnel Teaching-Related Infrastructure Issues Computational chemistry courses, interdisciplinary computational science courses, software engineering (not just programming), test problems and design case studies, support for software maintenance, modeling- and simulation-based courses; top two issues: modeling-and-simulation course, and mathematics training Research-Related Infrastructure Issues Multiscale modeling algorithms, representation and treatment of uncertain- ties, standard test cases (software, experiments), funding of interdisciplinary re- search, shared instruments; top two issues: multiscale modeling algorithms and funding of interdisciplinary research Equipment-Related Infrastructure Issues Access to high-performance computational facilities; high-bandwidth access to instruments and computers; collaborative environments; clusters and tightly
94 APPENDIX G coupled, diverse architectures; assimilation of information from sensors; interoperability and portability; distributed databases; shared resources (equip- ment, software, computers); access to a diverse set of architectures; top two is- sues: access to a diverse set of architectures, and interoperability and portability Codes and Software-Related Infrastructure Issues Integration of multivendor software systems, software maintenance (labora- tory technicians), open source, code and data warehouses, component architec- tures, interoperability, security; top two issues: component architectures and open source code Personnel Related-Infrastructure Issues Poor computer and software engineering skill levels of incoming students, tenure and promotion of interdisciplinary people, training of chemists with chemi- cal engineering concepts, people in the pipeline, attracting interest of computer science community; top two issues: tenure and promotion of inter-disciplinary people, and people in the pipeline Green-Red Group Report What Parts Are Working? Commercial software companies serving this area: computational chemistry, chemical processes Chemist-chemical engineer collaboration works well where it exists Modern programming tools have speeded code development Networking: Internet high-speed connectivity What Parts Are Not Working ? Commercial software efforts limited by market size: limited new science, porting to new machines, commercialization can kill technology, coding complex programs still very difficult, high-performance computing centers are not gener- ally useful, proposal process too restrictive, DOE centers working better than NSF centers Education of new computational chemists and chemical engineers inad- equate: undergraduate-graduate curriculum needs to address these subjects, ap- plications and scientific programming (Colorado School of Mines Engineering Web site is doing this) Academic code sharing and support mechanisms poor: code development
APPENDIX G 195 not supported by grants, no money or staff for maintenance and support, duplica- tion of effort on trivial code; need open source toolkits and libraries Databases and knowledge bases: not generally available except in bioinformatics, need standards and data validation Cross-discipline collaboration and teamwork inadequate: overspecialization in graduate school departments What Payoffs Are Expected ? Proof of profit demonstration that simulation methods can make difference in business outcome, "home run" in drug discovery, more incorporation of meth- ods into educational programs (better understanding of scope and limitations of methods), make specialists in industry more connected and useful (leveraging current staff), better communication and teaming of chemists and chemical engi- neers (more joint appointments, maybe combining departments), commonality in understanding and language between users and nonusers How Can Community Assure Reliability, Availability, and Maintenance of Codes ? Recognize contributions of coders, support full life cycle of codes, agencies should fund code development and maintenance as part of grants, demonstrate value of simulation codes, funding incentives for cross-disciplinary software creation
