An often-cited quote by the statistician George Box is that, “Essentially all models are wrong, but some are useful” [1]. Many modeling approaches for biology have been and are being developed with the aim of being very useful, and this paper describes one such approach.

Biocharts are based on David Harel’s influential notation called statecharts, which capture the concurrent evolution of various communicating processes, via the formalism known as live sequence charts, which attempt to capture the dynamics of systems with a strong notion of liveness. In adapting statecharts to biological domains, Kugler adds an object-oriented framework, arguing that this captures the notion of a population of cells in different states of execution. This notation and its computational basis are described in sections 2 and 3 of the paper.

Section 4 introduces some domains of application. The first one is chemotaxis; however, Kugler does not introduce any new question or hypotheses, but models the widely accepted idea that chemotaxis is a result of recognition of temporal variation rather than direct recognition of spatial gradients. The second example relates to the distribution of germlines in the development of the C. elegans worm, where different rates of cell division in different locations of an embryo are needed to ensure proper growth. The third example is related to this second one, where now a study is made to capture the lineage of cells during later stages in the life of the worm. All three case studies are classic examples in biology.

Developers of computer modeling systems ought to heed Box’s quote on usefulness. While the semantics of biocharts may be precise and powerful in capturing biological concepts, the proof of the pudding, so to speak, is in how useful the system could be in answering significant questions. In this paper, as unfortunately in many other attempts to provide a computational framework for biology, the main result is the representation of already known results. While this approach is reasonable considering that the paper was published under the aegis of the IEEE, I still find it disappointing that there is little that can provide experimentalists with any new insights into the case studies. One is left wondering why a formal modeling system should be used.

The author provides some pointers toward real usefulness of the system: in section 1, he states that biocharts can lead “towards unifying hypotheses with models and experiments,” but the paper does not provide evidence of this very useful unification. There is also a statement that the introduced methodology allows “scalable simulation, visualization and analysis”; however, there is no evidence provided for this link. Figures 7 to 12 (apart from 10) show the spatial aspects of the behavior, but it doesn’t seem that these images depict a direct outcome from the biochart modeling. So, while the author acknowledges the factors that could increase the usefulness of the model, at least this paper does not provide enough evidence for this usefulness.

If I were a prepublication referee for this paper, I would have insisted on several improvements. Figure 1 introduces both the representation as well as an example, but there is no explanation; therefore, the author misses an opportunity to clearly explain the formalism. The figures that illustrate the case studies are not well integrated with the text or with each other. For example, figures 7 and 8 represent different levels of detail of the structure of a C. elegans worm; however, they use two different color schemes, which creates unnecessary confusion.

Some of the deficiencies I point to above are probably a consequence of the restrictions faced by the author, such as space limitation, and the intended audience for the presentation. The paper, however, shows the promise of the biocharts formalism in expressing some of the features of the very interesting dynamics of biological systems.