Preparing the Next Generation of Integrative Organismal Biologists

Abstract

Pursuing cutting edge questions in organismal biology in the future will require novel approaches for training the next generation of organismal biologists, including knowledge and use of systems-type modeling combined with integrative organismal biology. We link agendas recommending changes in science education and practice across three levels: Broadening the concept of organismal biology to promote modeling organisms as systems interacting with higher and lower organizational levels; enhancing undergraduate science education to improve applications of quantitative reasoning and modeling in the scientific process; and K-12 curricula based on Next Generation Science Standards emphasizing development and use of models in the context of explanatory science, solution design, and evaluating and communicating information. Out of each of these initiatives emerges an emphasis on routine use of models as tools for hypothesis testing and prediction. The question remains, however, what is the best approach for training the next generation of organismal biology students to facilitate their understanding and use of models? We address this question by proposing new ways of teaching and learning, including the development of interactive web-based modeling modules that lower barriers for scientists approaching this new way of imagining and conducting integrative organismal biology.

Fig. 1

Example of a hypothetical gene network. The interactions of individual genes and gene products (mRNAs and proteins) are needed to understand the functioning of a cell.

Fig. 2

Within organisms are a variety of nested interaction networks that include genetic networks, physiological performance, development, and other processes. Organisms develop phenotypes, have behaviors and form populations. They are found within habitats that then affect those internal processes, and result in feedbacks with other organisms within a habitat. Local environments and habitats are similarly nested within ecosystems that influence responses of local systems as well as feedbacks on the larger scale processes.

Fig. 3

Conceptual structure for a library of online models, designed to function as a community-level distributed resource to facilitate experiential learning and quantitative research by organismal biologists. The key elements are a searchable database index that organizes library entries to facilitate finding models, activities, and background relevant to a given organizational level, organism, conserved quantity, mechanism, or computational platform of resources for students. Other elements of this online resource would include executable models, activities, and “quick explainers” of topics, which are brief summaries of jargon terms, concepts, or methods referred to in the executable models or activities.

Fig. 4

Examples of graphics and other content presented to students by the embryo swimming model, as implemented within Jupyter notebooks. In particular, this montage shows screenshots of what a student sees when using this notebook—self-contained background and usage instructions, simple interfaces for selecting parameters and visually engaging presentation of numerical results—all of which facilitate active learning and acquisition of modeling skills by organismal biology students. Some text may be difficult to read without high magnification, but even at low magnification the contrast of the Jupyter notebook framework with traditional command line inputs and outputs is clear. Elements of this user interface for simulating swimming in shear and rotating flow are shown, including the textboxes for specifying flow (in this case, pure rotation) and other simulation parameters (duration, ciliary velocity, initial orientation). At the bottom is the model output, including visualizations of larval trajectory, position, and orientation (that are played as a movie in real time), and numerical metrics of position, orientation, and velocity. In this example, the larva is able to maintain an inclined but stable orientation as it is advected in the rotating flow.

Fig. 5

The Jupyter notebook interface for the embryo swimming model, showing tools provided to students enabling them to “design” early larval shapes by mimicking traits of extant species or inventing hypothetical ones. The notebook also presents a 3D rendering of the embryo morphology and, as an option for interested students, the numerical gridpoints used in the fluid dynamics calculations (not shown). This notebook enables students to investigate effects on swimming performance of specific changes in the size, morphology, and composition of early stage larvae. Students use this notebook in conjunction with the notebook in Fig. 4, to vary larval traits in the context of environmental variations (e.g., turbulent near-surface habitats vs. quiescent deep-water habitats). Working with this model helps students to understand scenarios under which swimming requirements impose benefits or constraints on larval size and shape, and to speculate about implications for life history evolution, species ranges, and other large time/space scale processes.

Fig. 6

Screenshots of graphics and numerical output presented to students by the cholera model from Koelle et al. (2005), as implemented within a Jupyter notebook. The top plot shows key results from a cholera simulation scenario: Time series of infected and susceptible individuals (note different axes) across rapid seasonal variations and decade-scale “climatic” changes roughly corresponding to El Niño Southern Oscillation fluctuations. After studying and understanding these time series with default parameters, students consider the underlying environmental and physiological mechanisms. The middle plot visualizes how immunity (K_i) decays over time in individuals who have recovered from a cholera infection, an intuitive graphical presentation of a concept from epidemiology (“immune period”) that students often find confusing and mathematically complex. The lower plots show how fluctuations due to precipitation variations on seasonal and climatic timescales impact total transmission rate (β_t), which corresponds to the environmental contribution to the basic reproductive number, . Students complete the exercise by manipulating these driving mechanisms, considering the potential effects of alternative mitigation strategies, such as increasing the immune period by improving general health or reducing transmission by improving sanitation.