Biology Reference
In-Depth Information
that, independently from alternating conditions, the spliceosome reaches critical assembly checkpoints
with different protein complements. Alternatively, this might extend our view to different spliceosomes
in dependence on different cellular or environmental conditions. The existence of a major and a minor
spliceosome [Will and L uhrmann, 2005] with different mRNA substrate specificities could support this
notion, but the major spliceosome as considered here suggests for itself a highly dynamic assembly
process. The flood of different mechanistic examples of individual and sometimes quite different inter-
mediate steps makes it clear that there is no “
one spliceosome
”. Hence, there can be no single model of
spliceosome assembly.
Although the current model represents a higher coverage of experimentally supported subsystems in
the early (E- and A-complex) in comparison to later assembly stages, it is tempting to hypothesize that
the number of different routes increases with the importance of the intermediate complex for the overall
assembly process.
A further purpose of this model was to demonstrate how knowledge gained by experiments on the
spliceosome can be transformed into formalized descriptions, which are suitable for computational
analysis. Biochemical reactions, describing all steps of spliceosome assembly in more or less detail
and having accumulated over the past two decades, were extracted and assigned to reactions that can be
used for structural modeling. Hereby, the use of standardized identifiers for protein factors will greatly
ease the combination of smaller models and their successive integration into larger systems. The power
of predictive modeling is increasing as more submodels can be integrated, covering more details of the
spliceosome assembly pathway.
Several requirements to future works on spliceosome analysis can be asserted from this work, ad-
dressing both, experimental biologists and computational scientists. First, experimental data should
immediately be stored in a structured pre-formated way, making use of existing formalisms and avoiding
unnecessary naming morphisms. Experimental data provides precious facts, which are necessary to
prepare subsequent
in silico
analyses. Second, theoretical and computational contributions can still be
improved in the supply of data collection tools as well as integrated pipelines for their global evaluation
and analysis. For example, text mining tools at the level of network design and statistical measures at
the level of substructure analysis (
e.g.
, T-invariants, MCTSs) can enhance the output of this modeling
approach.
Finally, one needs to keep in mind that structural properties depend at first hand on the knowledge put
into the model. In light of the wealth of biological information, a direct consequence is the possibility that
parts of the model are better covered by data than others, and therefore exhibiting a higher complexity.
Consequently, these parts are stronger represented by T-invariants. Nevertheless, the fact to observe
a stronger representation of individual aspects of a biological network justifies a model, because it
captures relations and trends that are hard to detect using detailed mechanistic studies, which moreover
are too numerous for individual analyses. The first mathematical model of spliceosomal assembly
pathway presented in this paper could serve as the basis for further investigations, experimental as well
as theoretical ones.
ACKNOWLEDGEMENTS
We want to thank the anonymous reviewers for their helpful comments on the manuscript. Financial
support from the German Ministry for Research and Education (BMBF) to RHB within the HepatoSys
framework is gratefully acknowledged.
