Optimizing agent-based transmission models for infectious diseases

BMC Bioinformatics

Lander Willem⁵  Sean Stijven⁴  Engelbert Tijskens³  Philippe Beutels²  Niel Hens⁵  Jan Broeckhove¹ 

[1] Modeling of Systems And Internet Communication, Department of Mathematics and Computer Science, University of Antwerp, Antwerp, Belgium

[2] School of Public Health and Community Medicine, The University of New South Wales, Sydney, Australia

[3] HPC core facility CalcUA, Computational Mathematics, University of Antwerp, Antwerp, Belgium

[4] Department of Information Technology, Ghent University–iMinds, Ghent, Belgium

[5] Interuniversity Institute for Biostatistics and statistical Bioinformatics, Hasselt University, Hasselt, Belgium

关键词: Performance;    Optimization;    Agent-based model;    Mathematical epidemiology;   

Others  :  1232217
DOI  :  10.1186/s12859-015-0612-2

 received in 2014-11-21, accepted in 2015-05-11,  发布年份 2015

【 摘 要 】

Background

Infectious disease modeling and computational power have evolved such that large-scale agent-based models (ABMs) have become feasible. However, the increasing hardware complexity requires adapted software designs to achieve the full potential of current high-performance workstations.

Results

We have found large performance differences with a discrete-time ABM for close-contact disease transmission due to data locality. Sorting the population according to the social contact clusters reduced simulation time by a factor of two. Data locality and model performance can also be improved by storing person attributes separately instead of using person objects. Next, decreasing the number of operations by sorting people by health status before processing disease transmission has also a large impact on model performance. Depending of the clinical attack rate, target population and computer hardware, the introduction of the sort phase decreased the run time from 26 % up to more than 70 %. We have investigated the application of parallel programming techniques and found that the speedup is significant but it drops quickly with the number of cores. We observed that the effect of scheduling and workload chunk size is model specific and can make a large difference.

Conclusions

Investment in performance optimization of ABM simulator code can lead to significant run time reductions. The key steps are straightforward: the data structure for the population and sorting people on health status before effecting disease propagation. We believe these conclusions to be valid for a wide range of infectious disease ABMs. We recommend that future studies evaluate the impact of data management, algorithmic procedures and parallelization on model performance.

【 授权许可】

   
2015 Willem et al.; licensee BioMed Central.

【 图 表 】

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

【 参考文献 】

• [1]Eubank S, Guclu H, Kumar VSA, Marathe MV, Srinivasan A, Toroczkai Z et al.. Modelling disease outbreaks in realistic urban social networks. Nature. 2004; 429(6988):180-4.
• [2]Ferguson NM, Cummings DAT, Cauchemez S, Fraser C, Riley S, Meeyai A et al.. Strategies for containing an emerging influenza pandemic in Southeast Asia. Nature. 2005; 437(7056):209-14.
• [3]Germann TC, Kadau K, Longini Jr IM, Macken CA. Mitigation strategies for pandemic influenza in the United States. Proc Natl Acad Sci. 2006; 103(15):5935-40.
• [4]Degli Atti MLC, Merler S, Rizzo C, Ajelli M, Massari M, Manfredi P et al.. Mitigation measures for pandemic influenza in Italy: an individual based model considering different scenarios. PLoS ONE. 2008; 3(3):1790.
• [5]Das TK, Savachkin AA, Zhu Y. A large-scale simulation model of pandemic influenza outbreaks for development of dynamic mitigation strategies. IIE Trans. 2008; 40(9):893-905.
• [6]Chao DL, Halloran ME, Obenchain VJ, Longini Jr IM. FluTE, a publicly available stochastic influenza epidemic simulation model. PLoS Comp Biol. 2010; 6(1):1000656.
• [7]Roche B, Drake JM, Rohani P. An agent-based model to study the epidemiological and evolutionary dynamics of influenza viruses. BMC Bioinformatics. 2011; 12(1):87. BioMed Central Full Text
• [8]Laskowski M, Demianyk BC, Witt J, Mukhi SN, Friesen MR, McLeod RD. Agent-based modeling of the spread of influenza-like illness in an emergency department: a simulation study. IEEE Trans Inf Technol Biomed. 2011; 15(6):877-89.
• [9]Aleman DM, Wibisono TG, Schwartz B. A nonhomogeneous agent-based simulation approach to modeling the spread of disease in a pandemic outbreak. Interfaces. 2011; 41(3):301-15.
• [10]Grefenstette JJ, Brown ST, Rosenfeld R, DePasse J, Stone NT, Cooley PC et al.. FRED (A Framework for Reconstructing Epidemic Dynamics): an open-source software system for modeling infectious diseases and control strategies using census-based populations. BMC Public Health. 2013; 13(1):940. BioMed Central Full Text
• [11]Halloran ME, Ferguson NM, Eubank S, Longini IM, Cummings DA, Lewis B et al.. Modeling targeted layered containment of an influenza pandemic in the United States. Proc Natl Acad Sci. 2008; 105(12):4639-644.
• [12]Willem L, Stijven S, Vladislavleva E, Broeckhove J, Beutels P, Hens N. Active learning to understand infectious disease models and improve policy making. PLoS Comput Biol. 2014; 10(4):1003563.
• [13]Drepper U. What every programmer should know about memory. http://www.akkadia.org/drepper/cpumemory.pdf.
• [14]Giles M, Reguly I. Trends in high-performance computing for engineering calculations. Phil Trans R Soc A. 2014; 372(2022):20130319.
• [15]Giles MB, Mudalige GR, Sharif Z, Markall G, Kelly PH. Performance analysis and optimization of the OP2 framework on many-core architectures. SIGMETRICS Perform. Eval. Rev. 2011; 38:9-15.
• [16]Sutter H. The free lunch is over: A fundamental turn toward concurrency in software. Dr. Dobb s J. 2005; 30(3):202-10.
• [17]Sutter H, Larus J. Software and the concurrency revolution. Queue. 2005; 3(7):54-62.
• [18]Hayes B. Randomness as a resource. Am Sci. 2001; 89(4):300-4.
• [19]Bauke H, Mertens S. Random numbers for large-scale distributed Monte Carlo simulations. Phys Rev E. 2007; 75(6):066701.
• [20]Mossong J, Hens N, Jit M, Beutels P, Auranen K, Mikolajczyk R et al.. Social contacts and mixing patterns relevant to the spread of infectious diseases. PLoS Med. 2008; 5(3):74.
• [21]Hens N, Ayele GM, Goeyvaerts N, Aerts M, Mossong J, Edmunds JW et al.. Estimating the impact of school closure on social mixing behaviour and the transmission of close contact infections in eight European countries. BMC Infect Dis. 2009; 9(1):187. BioMed Central Full Text
• [22]Willem L, Van Kerckhove K, Chao DL, Hens N, Beutels P. A nice day for an infection? Weather conditions and social contact patterns relevant to influenza transmission. PloS ONE. 2012; 7(11):48695.
• [23]RTI International: 2010 RTIU.S. synthetic population ver. 1.0. 2014 Downloaded from http://www.epimodels.org.
• [24]Wheaton W. 2010 U.S. synthetic population quick start guide. RTI international. 2014. Retrieved from http://www.epimodels.org.
• [25]Chapman B, Jost G, Van Der Pas R. Using OpenMP: Portable Shared Memory Parallel Programming vol. 10. MIT Press, Massachusetts, USA; 2008.
• [26]Bauke H, Brown WE, Fischler M, Kowalkowski J, Paterno M, Knuth DE, Press WH, Teukolsky SA, Vetterling WT, Flannery BP, et al. Tina‘s random number generator library. 2011. Retrieved from http://numbercrunch.de/trng/.
• [27]Google: C++ testing framework. 2014. Retrived from http://code.google.com/p/googletest/.
• [28]Smart J. Jenkins: the definitive guide. O’Reilly Media, Sebastopol, California; 2011.
• [29]Tclap. The templetized C++ command line parser library. 2014. Retrieved from http://www.tclap.sourceforge.net.
• [30]Fialho L, Browne J. Framework and modular infrastructure for automation of architectural adaptation and performance optimization for HPC systems. Supercomputing. Springer, Cham, Switzerland; 2014.
• [31]Rane A, Browne J. Enhancing performance optimization of multicore/multichip nodes with data structure metrics. ACM Trans Par Comput. 2014; 1(1):3.