2 Background, Tasks, Modeling Methods …
29
2.4.2 In the Face of Complex Living Systems
Although the fugacity models possess simplified structures under the principle of
Ockham’s Razor, they are indeed good at describing the fate of chemicals within
an environmental system if properly parameterized [31]. The same goes for PBTK
models. As mentioned previously, these models neglect the counter-influence of
chemicals on respective situational objects, simulating merely the fate of chemicals
rather than the fate of the whole systems, and thus are relatively straightforward to
build and to simulate.
However, at cellular or subcellular levels, things are significantly different.
Because it is the fates of the biological systems that are focused on in the simulations, the counter-influence of chemicals on the cellular components and function
cannot be ignored any more. To enable the simulation of perturbation on biological
systems from chemicals, mechanisms and functional structures of biological systems themselves and interfaces between the biological systems and the xenobiotic
chemicals must be firstly known. It is well acknowledged that modern toxicology
studies based on in vitro assays have successfully parsed particular interfaces. However, decoding the mystery of life is never an easy job. The functional structures of
most biological systems are still not clear. Moreover, there is a lack of knowledge
on relevant gene polymorphisms of human populations. Computational toxicologists
obviously cannot digitize so many objects within these unknown fields. This state
will hopefully be improved with systems biology and “-omics” technologies that are
emphasized to piece the whole picture of life together [46, 47].
Furthermore, the ultimate goal of chemicals risk assessment is to protect not only
human beings, but also the diverse species that inhabit the earth’s ecosystems, which
is also a topic of ecotoxicology [85]. In fact, the sensitivity of different species
to the same chemicals could vary distinctly. However, it is neither pragmatic nor
necessary to evaluate the toxic effects of chemicals on more than a million species
via wet experiments. A key point is to extrapolate toxicities of chemicals cross
different species, which also requires a sound understanding of mechanisms and
functional structures of different concerned species [86]. Computational toxicology
has provided some strategies to address the problem of cross-species extrapolation.
For example, by modifying structures and parameters of PBTK models, internal
distribution or dynamic bioaccumulation of chemicals across different species can
be evaluated [87]. To simulate biomacromolecules from different species, homology
modeling [88] could provide atom-level structures that have not been determined
by X-ray diffraction or nuclear magnetic resonance approaches. However, there is
still very little knowledge of non-human species, except for a very limited number
of model species [89], which can be employed for establishing in silico models.
Therefore, feasible and convincing models for cross-species extrapolation are still
far from real practice.
29
2.4.2 In the Face of Complex Living Systems
Although the fugacity models possess simplified structures under the principle of
Ockham’s Razor, they are indeed good at describing the fate of chemicals within
an environmental system if properly parameterized [31]. The same goes for PBTK
models. As mentioned previously, these models neglect the counter-influence of
chemicals on respective situational objects, simulating merely the fate of chemicals
rather than the fate of the whole systems, and thus are relatively straightforward to
build and to simulate.
However, at cellular or subcellular levels, things are significantly different.
Because it is the fates of the biological systems that are focused on in the simulations, the counter-influence of chemicals on the cellular components and function
cannot be ignored any more. To enable the simulation of perturbation on biological
systems from chemicals, mechanisms and functional structures of biological systems themselves and interfaces between the biological systems and the xenobiotic
chemicals must be firstly known. It is well acknowledged that modern toxicology
studies based on in vitro assays have successfully parsed particular interfaces. However, decoding the mystery of life is never an easy job. The functional structures of
most biological systems are still not clear. Moreover, there is a lack of knowledge
on relevant gene polymorphisms of human populations. Computational toxicologists
obviously cannot digitize so many objects within these unknown fields. This state
will hopefully be improved with systems biology and “-omics” technologies that are
emphasized to piece the whole picture of life together [46, 47].
Furthermore, the ultimate goal of chemicals risk assessment is to protect not only
human beings, but also the diverse species that inhabit the earth’s ecosystems, which
is also a topic of ecotoxicology [85]. In fact, the sensitivity of different species
to the same chemicals could vary distinctly. However, it is neither pragmatic nor
necessary to evaluate the toxic effects of chemicals on more than a million species
via wet experiments. A key point is to extrapolate toxicities of chemicals cross
different species, which also requires a sound understanding of mechanisms and
functional structures of different concerned species [86]. Computational toxicology
has provided some strategies to address the problem of cross-species extrapolation.
For example, by modifying structures and parameters of PBTK models, internal
distribution or dynamic bioaccumulation of chemicals across different species can
be evaluated [87]. To simulate biomacromolecules from different species, homology
modeling [88] could provide atom-level structures that have not been determined
by X-ray diffraction or nuclear magnetic resonance approaches. However, there is
still very little knowledge of non-human species, except for a very limited number
of model species [89], which can be employed for establishing in silico models.
Therefore, feasible and convincing models for cross-species extrapolation are still
far from real practice.