Abstract
Scientific practices have been changed by the increasing use of computer simulations. A central question for philosophers is how to characterize computer simulations. In this paper, we address this question by analyzing simulations in biochemistry. We propose that simulations have been used in biochemistry long before computers arrived. Simulation can be described as a surrogate relationship between models. Moreover, a simulative aspect is implicit in the classical dichotomy between in vivo–in vitro conditions. Based on a discussion about how to characterize a simulative aspect in this dichotomy, we will argue that an adequate understanding of computer simulations in biochemistry requires a previous understanding of simulations in experimental contexts.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
This approach can be traced back to Humphrey’s concept of computational science Humphreys (2004).
We want to highlight that our analysis is not, in principle, restricted to the particular examples considered here; it may also be applied to fields that share similar methodological resources.
There are other goals in philosophical accounts of computer simulations like methodological novelty or taxonomies. But the usual aim is to obtain a general concept of simulation. In a recent paper, Juan Duran present an account of the philosophy of computer simulations (Duran 2013).
From this perspective, when a system can replace, to same extent, another system, we can talk about an imitative relationship.
Of course, in vitro simulations are sometimes used to overcome ethical problems, but here we want to stress only the methodological role of simulations cfr. Kilkenny et al. (2010) p. 186ss.
We do not need to rest on an account of resemblance or similarity. We only need to describe the methodological techniques that allows us to construct a surrogate system.
In this paper we do not try to trace the history of in vivo- in vitro techniques. We only want to point out, through some examples, how the methodological concern behind the in vivo- in vitro dichotomy is present since the origins of biochemistry and related fields.
However perfusion methods were never completely abandoned. We can read scientific reports that still use this technique (Cockcroft et al. 2009).
If we are interested in intermediate metabolism, we need to get some resources for investigating something more than the input of the system or their final result (an input–output schema).
In short: an experimental simulation—in vitro setup—of in vivo conditions is possible because of the tension between imitative and disruptive aspects. And, as we said above, we understand the imitative aspect in terms of the techniques that enable a certain degree of confidence in the surrogate system.
This last point can be reinforced by taking into account in vivo models and their byproducts. In general these in vivo models refer to animal models. The expression "animal model" it is not only used to refer to the use of animals in experimenting, but also for diverse ways and levels of intervention, including genetic modification of specimens to make the model more similar to the system under study. For example, NSG mice (also called "humanized mice models") are used in several laboratories. It is worth noticing that an improvement in the understanding of biological mechanisms has as a consequence the substitution of traditional animal models by tissue probes. Because of their construction assumptions, techniques and functions, those experimental arrays are called ‘models’. But these assumptions, techniques and function are derived from in vivo or in vitro conditions. Therefore, to consider model-building context far from being an objection can be seen as a way to reinforce our main thesis.
is we can highlight this statement by pointing out the sources and levels of the disruptive aspects. There is a simplification on a general level, due to the generic model building process; there is also a disruption due to a comparison between in vitro—tissue slices—and in vivo—experimental results from an input–output schema—conditions. Both conditions are disruptive aspects but with a different source, and only the second one is directly relevant for our notion of simulation.
By speaking of explicit and implicit aspects of computer program building process, we only want to make a distinction between different levels. One of them in which there is an explicit process for constructing a scientific model. And the other level, represented by implicit assumptions that are a fixed part (more than the input data) of the computer program that runs the simulation.
However, modelization in analog devices has some philosophical interest. As Mendes points out, those models were not considered as a simplification of the modeled situation (Mendes 1994, p. 126–128). In those cases the imitative aspect of modeling is considered almost a mimicry. Mendes is not defending a positions here, but only reporting the usual believes of scientist at that time (Mendes 1994). However, it seems that Mendes considerations are based in two aspects that are not clearly distinguished by him. In the cited Chance’s (1943) paper, there are two devices: an experimental array—a stop-flow device, a device for studying the kinetics of a compound- in order to determining enzyme mechanism and a digital analyzer for integrating the ODEs (Chance 1943). Mendes considers the first type of experimental array as a simulation, because it is a device that allows to ‘imitate’ enzyme behavior. The dynamics of the physical device—mechanical, electrical or even hydraulic- are responsible for the simulative task mediated by the numerical solution.
The entire project behind Gepasi has been reformulated by Pedro Mendes with a program for simulating biochemical networks called Copasi—COmplex PAthway Simulator-. (More information could be found at http://www.copasi.org).
18METAMOD is a software package to study steady-state metabolic pathways. This package also allows the study of control analysis of metabolic pathways. Scamp is a general-purpose simulator of metabolic reaction.
For this particular research, METAMOD is used (Cornish-Bowden 1991). But, the same results could be obtained with Gepasi.
Pedro Mendes questioned the Cornish-Bowden’s conclusion in several articles. Mendes showed how the channeling process could decrease free metabolites concentration by means of some parameter variations. Then, the discussion centered on the legitimacy of changing the parameters in biochemical simulations. Nevertheless, from those simulations, it is clear that changing the parameters in the model could affect the reversibility of the channeling reaction. And the reversibility of channeling could be a reason why the concentration did not decrease, because a reversible reaction could “inhibit the ability of any channel to decrease a pool” (Mendes et al. 1992, p. 259). In this context “pool” refers to a metabolite concentration.
Channeling could occur in static complexes of enzymes or in dynamic complexes. The term ‘dynamic’ in this last case came from the short life of enzymes in the complexes (Mendes 1994, pp. 12–13).
According to Winsberg, if we want to have a computational model of some phenomena, we usually have to follow some steps that can be described in terms of models. The first kind of resource that allows us to work with a theory is called for Winsberg a “mechanical model”: “a mechanical model is a bare bones characterization of a physical system that allows us to use the theoretical structure to assign a family of equations to the system.” (Winsberg 1999). In the case of a program like Gepasi, when a reaction model is define -for example an irreversible reaction- the corresponding differential equations are generated. When we restrict this model to a class of phenomena with the specification of parameters, boundary and initial conditions, we have a “dynamic model”. In Gepasi we can specify the kinetic scheme or choose a reaction mechanism like a Michaelis–Menten enzyme kinetics. Then the user has to define the parameters of the reaction. We describe the setup of the simulation from the point of view of the final user. If we change the perspective to the builder of the simulation, restrictions about what is tractable appear. Winsberg point out that sometimes some “ad hoc” adjustments are needed (Winsberg 2010). This step is very important for the program builder, but in the case of Gepasi it is “invisible” to the user of the program.
This particular consideration can be understood in terms of Humprheys’ templates concept. In a nutshell, a template is a mathematical structure that can be used for modeling in different disciplines (Humphreys 2004).
There are other main types like reversible or irreversible inhibitions and there is also a division between noncompetitive and uncompetitive inhibitions. But, here we are only concerned with competitive and uncompetitive inhibitions (Leskovac 2003, p. 73).
According to Cornish-Bowden, those “stoichiometric constrains” rest mainly on “algebra”, because he uses Gaussian elimination to analyze networks that represent stoichiometric relationship (Cornish-Bowden and Hofmeyr 2002).
This target is also interesting because some experimental observation had shown that inhibiting it in vivo bring about a concentration rise.
The perspective of MCA requires some knowledge about in vivo kinetic properties of enzymes. And most of the currently available information is from in vitro experiments (Wang et al. 2004).
MCA made very different assumptions about irreversible or reversible reaction when it uses technical concepts like “elasticity”. This is a property of an enzyme that consists in the relationship of the variation of the rate regarding the concentration of a metabolite (Cornish-Bowden and Eisenthal 2000).
References
Chance, B.: The kinetics of the enzyme-substrate compound of peroxidase. J. Biol. Chem. 151, 553–577 (1943)
Cockcroft, N.Y., Oke, O., Cunningham, F., Bishop, E., Fearon, I.M., Zantl, R., Gaça, M.D.: An in vitro perfusion system to examine the responses of endothelial cells to simulated flow and inflammatory stimulation. Altern. Lab. Anim. ATLA 37(6), 657–669 (2009). (PMID:20105001)
Cornish-Bowden, A.: Failure of channelling to maintain low concentrations of metabolic intermediates. Eur. J. Biochem. 195(1), 103–108 (1991)
Cornish-Bowden, A., Cardenas, M.L.: Metabolic analysis in drug design. C R Biol 326(5), 509–515 (2003)
Cornish-Bowden, A., Cárdenas, M.L., Letelier, J.-C., Soto-Andrade, J., Abarzúa, F.G.: Understanding the parts in terms of the whole. Biol. Cell 96(9), 713–717 (2004). doi:10.1016/j.biolcel.2004.06.006
Cornish-Bowden, A., Eisenthal, R.: Computer simulation as a tool for studying metabolism and drug design. In: Cornish-Bowden, A., Cárdenas, M. (eds.) Technological and Medical Implications of Metabolic Control Analysis, vol. 74, pp. 165–172. Springer, Berlin (2000)
Cornish-Bowden, A., Hofmeyr, J.-H.S.: The role of stoichiometric analysis in studies of metabolism: an example. J. Theor. Biol. 216(2), 179–191 (2002). doi:10.1006/jtbi.2002.2547
Decock, L., Douven, I.: Similarity after Goodman. Rev. Philos. Psychol. 2(1), 61–75 (2010)
Duran, J.: A brief overview of the philosophical study of computer simulations. APA Newsl. 1(13), 38–46 (2013)
Eccles, S.A.: Basic principles for the study of metastasis using animal models. In: Brooks, S., Schumacher, U. (eds.) Metastasis Research Protocols, pp. 161–171. Springer (2001)
El Skaf, R., Imbert, C.: Unfolding in the empirical sciences: experiments, thought experiments and computer simulations. Synthese 190(16), 3451–3474 (2013). doi:10.1007/s11229-012-0203-y
Fruton, J.: The emergence of biochemistry. Science 192, 327–334 (1973)
Fruton, J.: Proteins, Enzymes, Genes: The Interplay of Chemistry and Biology. Yale University Press, New Haven (1999)
Gilbert, G.N., Troitzsch, K.G.: Simulation for the social scientist. In: Hill, M. (ed.). Open University Press. http://public.eblib.com/EBLPublic/PublicView.do?ptiID=287870 (2005). Retrieved 22 March 2014
Goodman, N.: Languages of Art: An Approach to a Theory of Symbols. Hackett, Indianapolis (1976)
Guala, F.: Models, simulations, and experiments. In: Magnani, L., Nersessian, N. (eds.) Model-Based Reasoning, pp. 59–74. Springer, New York (2002)
Hartmann, S.: The world as a process: simulations in the natural and social sciences. In: Hegselmann, R., et al. (eds.) Modelling and Simulation in the Social Sciences from the Philosophy of Science Point of View, pp. 77–100. Kluwer, Dordrecht (1996)
Haunschild, M.D., Freisleben, B., Takors, R., Wiechert, W.: Investigating the dynamic behavior of biochemical networks using model families. Bioinformatics 21(8), 1617–1625 (2005)
Holmes, F.L.: Hans Krebs. Oxford University Press, New York (1991)
Holmes, F.L.L.: Hans krebs and the discovery of the ornithine cycle. Fed. Proc. 39(2), 216–225 (1980)
Humphreys, P.: Computer simulations. In: Psa: Proceedings of the Biennial Meeting of the Philosophy of Science Association, pp. 497506, (1990)
Humphreys, P.: Extending Ourselves: Computational Science, Empiricism, and Scientific Method. Oxford University Press, New York (2004)
Kilkenny, C., Browne, W., Cuthill, I.C., Emerson, M., Altman, D.G.: Animal research: reporting in vivo experiments: the ARRIVE guidelines. British J. Pharmacol. 160(7), 1577–1579 (2010)
Kohler, R.E. Jr.: The enzyme theory and the origin of biochemistry. Isis 64, 181–196 (1973)
Leskovac, V.: Comprehensive Enzyme Kinetics. Kluwer Academic/Plenum Pub, New York (2003)
Mendes, P.: Computer Simulation of the Dynamics of Biochemical Pathways (Unpublished doctoral dissertation). University of Wales, Aberystwyth (1994)
Mendes, P., Kell, D.B., Welch, G.R.: Metabolic channeling in organized enzyme systems: experiments and models. Adv. Mol. Cell Biol. 11, 119 (1995)
Mendes, P., Kell, D.B., Westerhoff, H.V.: Channelling can decrease pool size. Eur. J. Biochem. 204(1), 257–266 (1992)
Morgan, M.S.: Experiments versus models: new phenomena, inference and surprise. J. Econ. Methodol. 12(2), 317–329 (2005)
Morgan, J.L., Darling, A.E., Eisen, J.A.: Metagenomic sequencing of an in vitro-simulated microbial community. PLoS One 5(4), e10209 (2010)
Nielsen, J.: Metabolic control analysis of biochemical pathways based on a thermokinetic description of reaction rates. Biochem. J. 321, 133–138 (1997)
Parker, W.S.: Does matter really matter? Computer simulations, experiments, and materiality. Synthese, 169 (3), 483–496 (2009). Retrieved from. http://link.springer.com/10.1007/s11229-008-9434-3
Rohrlich, F: Computer simulation in the physical sciences. In: Proceedings of the Biennial Meeting of the Philosophy of Science Association, (1990)
Rostami-Hodjegan, A., Tucker, G.T.: Simulation and prediction of in vivo drug metabolism in human populations from in vitro data. Nat. Rev. Drug Discov. 6(2), 140–148 (2007)
Strand, R., Fjelland, R., Flatmark, T.: In vivo interpretation of in vitro effect studies with a detailed analysis of the method of in vitro transcription in isolated cell nuclei. Acta. Biotheor. 44(1), 121 (1996)
Tomita, M., Hashimoto, K., Takahashi, K., Shimizu, T.S., Matsuzaki, Y., Miyoshi, F.: E-cell: software environment for whole-cell simulation. Bioinformatics 15(1), 72–84 (1999)
Wang, L., Birol, I., Hatzimanikatis, V.: Metabolic control analysis under uncertainty: framework development and case studies. Biophys. J. 87(6), 3750–3763 (2004)
Winsberg, E.: Sanctioning models: the epistemology of simulation. Sci. Context 12(2), 275–292 (1999)
Winsberg, E.: Simulated experiments: methodology for a virtual world. Philos. Sci. 70(1), 105–125 (2003)
Winsberg, E.: A tale of two methods. Synthese 169(3), 575–592 (2009)
Winsberg, E.: Science in the Age of Computer Simulation. University of Chicago Press, Chicago (2010)
Acknowledgments
I would also like to thank Agencia Nacional de Promocion Cientìfica for the research grant (dir. Victor Rodriguez) that supported the work of this paper, as well as Secyt (Universidad Nacional de Cordoba).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Garcia, P. Computer simulations and experiments: in vivo–in vitro conditions in biochemistry. Found Chem 17, 49–65 (2015). https://doi.org/10.1007/s10698-015-9215-2
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10698-015-9215-2