World Wide Web Based Simulations for Teaching Biology

By Jeffrey R. Bell

 

The Biology Lab On-Line Project is a component of the California State University System (CSU) Integrated Technology Strategy (ITS), which calls for anywhere, anytime access to information. The project initially brought together biologists from throughout the CSU system and the CSU Center for Distributed Learning (CDL) to explore ways to use technology to improve learning in introductory biology laboratories. Later, multimedia developers from Addison Wesley Longman were added to the development team. A major goal of the collaboration was to allow students to learn as biologists do, i.e., by actively designing experiments and interpreting their results. Another goal was to extend student learning opportunities by creating simulations of experiments that they might not normally do because of expensive and/or inaccessible lab equipment, lack of field opportunities, time constraints, complexity, danger, or ethical problems. Eliminating the time constraints of the traditional experiment and two or three hour laboratory period gives students the opportunity to design and interpret experiments, learn from their mistakes, and to revise and redo their experiments just like real scientists. The simulations are not designed to replace the traditional "wet labs" found in the normal biology course, but rather are designed to extend the laboratory experience to subjects and experiments that can not normally be done, or not done enough, in a traditional laboratory. The simulations are also not multi-media presentations, stand-alone tutorials or on-line courses.

The project has produced five different educational simulations covering the subjects of evolution, Mendelian genetics, protein translation, human population demography, and protein structure-function. Current projects in progress and due to be finished by the summer of 1999 will simulate human genetics, mitochondrial electron transport, glycolysis, human female reproductive physiology, and photosynthesis. While each of the simulations is unique, all of them share many common interface elements and functions. All of the simulations have been designed so that the student can carry out many different experiments, allowing the student to design and interpret their own experiments. The flexibility of the programs make well designed exercises an important part of each laboratory. However, while sample exercises will be included with each lab, instructors can easily design their own exercises to meet the needs of their students.

The simulations have all been created in the Java programming language, so that they can be easily accessed over the web using any standard browser. This solves the problem of widely disseminating the applications, a common problem with most educational software. The Java application provides the user interface where students set the starting parameters for their experiment and get graphical feedback on their current settings. In some of the simulations the Java program also calculates the results, while in others the input parameters are passed back to the server, where the real calculations take place. When the server is done it sends the results back to the Java application, which presents the results to the student.

The downside of using Java is that only individuals and schools with fairly new computers and software (Netscape 3 or better, etc.) and an internet connection can use the software. Another disadvantage of using Java is the inability of Java programs to save to disk or print. This limitation has been overcome through the use of a notebook that can be exported to a web page. All of the data tables, such as numbers of different types of progeny, or results of statistical calculations, can be imported directly into the notebook. After typing in their comments the student can export the notebook to a web page for printing, or to email to themselves or an instructor. The web page is temporarily stored on the server. Graphical images such as graphs and charts produced by some of the programs are also exportable to the notebook, where they can then be printed.

All of the programs share some common user interface elements, including a title bar with links to an introduction to the lab, help, sample assignments, the notebook, etc. While there is much diversity in how the different labs operate, most of them start in an input mode where the student designs their experiment by adjusting different parameters. After designing the experiment the student runs the simulation. The program calculates the results of the experiment, usually in a minute or less, and then presents the results in the output mode. In this mode there is a tabbed interface where the student chooses which type of output they wish to view, a table of the data, a graph, the input values, etc. After analyzing their results they can import them into the notebook and then go back to the input mode to design another experiment. This ability to quickly go back and forth between the design of an experiment and the results is one of the powerful advantages of a simulation approach to teaching science.

Five of the programs are currently available for beta testing and all ten should be finished by the summer of 1999. Descriptions of each of the labs can be found below. Current plans call for a $19.95 fee for access to all ten of the simulations and a lab manual with printed instructions and sample assignments. The fee is necessary to support the servers and for maintenance of the various programs as operating systems and computers change. Below is a fairly complete description of EvolveIT along with brief descriptions of the other simulations. However, the best way to learn about these simulations is to go to the CDL site, http://www.cdl.edu/html/biology.html, and try out some of the five simulations currently available there.

EvolveIT

Although evolution is the unifying theme of the biological sciences, it is perhaps one of the most misunderstood and difficult concepts to convey in a laboratory setting. The study of evolution is especially suited to computer simulations because evolution normally occurs over very long time intervals, large data sets are usually needed to understand it, and there are usually a number of important parameters that are difficult to control in real experiments. EvolveIT (http://www.cdl.edu/EvolveIT), is a web based, interactive computer simulation designed to teach the basic concepts of natural selection and to convey the importance of time in the evolutionary process.

Students using EvolveIT observe evolutionary changes in bird beak morphology in hypothetical populations of birds isolated on two islands. In the simulation students can set the annual rainfall on island(s) containing finch populations, and then observe the effect of this environment on the evolution of the finch's beaks. Students may also change several other properties of the bird populations, such as initial mean beak size, beak size variability, beak size heritability and mean clutch size, to determine their effect on beak evolution.

To help students clearly see the effects of changes in the different variables, the simulation uses two islands with independent and isolated populations. A student can either directly compare two different sets of conditions, or do a duplicate run where both populations have identical starting conditions. As many students will also be unfamiliar with some of the variables, the program gives extensive feedback on what is being changed with each alteration of one of the initial variables. For instance, when the slider for mean beak size is moved to the right a picture of a finch head shows the finch beak growing larger, or when variability is increased a graph of the current distribution of beak sizes spreads out.

The program creates several hundred different virtual "birds" using the initial parameters entered by the student. These birds then go through a round of natural selection where each bird's probability of survival is a function of the seed distribution (determined by the setting for rainfall) and that bird's beak size. After the selection step the surviving birds are randomly mated to one another. They then produce offspring based on the values for heritability and variability entered by the student. This new population of birds becomes the starting population for the next year. This repeats for each year of the simulation. Because of the random selection and mating each simulation run is unique and will produce a different result than any other run, even one that starts with exactly the same starting parameters. The program produces several different outputs: a scrolling table with the mean beak size, the variability and the population size for every year of the simulation, for both populations; a series of histograms showing what proportion of birds of different beak sizes survived, for each year of the simulation; a plot of the mean beak size versus time for both populations; and a plot of the population size versus time for both populations. The final output is a table with the initial values for the simulation.

The student can rerun the simulation with the same initial values, revise the experiment, or start over with the default values and design a new experiment. Students can study the effects of different amounts of rainfall on the evolution of beak size to get a feel for how natural selection works. They should be able to determine that large beaks are favored in low rainfall, the optimum beak size for different amounts of rainfall, the effect of varying the severity of the selection, and the importance of the environment in determining the direction of selection. Similar experiments can be done in which only population variability or beak size heritability is manipulated to study how these parameters affect beak evolution. Changes in island size affect the carrying capacity of the island and allow the student to investigate the stochastic effects that can result from small population size. Variations in clutch size permit the student to investigate the consequences of different fecundities on the capacity of different species to evolve. Some aspects of population dynamics can also be investigated using this parameter (rate of exponential growth, boom bust population cycles, etc.)

While the simulation is based on Darwin's finches, changes in the species variables such as mean beak size, variability, heritability and clutch size create virtual species that can have properties similar to many other wild species. Students can investigate the parameters that are more likely to lead to the extinction of endangered species, see why some species might evolve faster than others, and examine many other facets of evolution. The program generates large data sets, one run can produce 600 data points, so students can learn how to analyze and interpret large amounts of data, unlike the situation in a typical lab. The great flexibility of the program should allow individual instructors to tailor student assignments to their particular preferences and provide students with a real opportunity to design their own experiments. Actively engaging students in exploring and studying evolution through this simulation will provide another avenue for students to learn about evolution in addition to the traditional text and lecture explanations.

Virtual Fly Lab

Two genetics labs are currently planned. One is a simulation of classical fruit fly genetics while the other one lets the student study human genetics by analyzing pedigrees. The Virtual Fly Lab (http://www.cdl.edu/Flylab) simulation is an update to the Virtual Fly Lab originally created by Bob Desharnais. In the Virtual Fly Lab students design their own fruit flies by choosing from many different possible phenotypes for characteristics such as eye color, wing shape, body color, etc. They then mate their flies and analyze the progeny to determine the rules of inheritance for different traits. Each experiment is unique and students can have up to 10,000 progeny produced from one mating. Offspring can also be mated so a wide range of different experiments are possible. There are 29 different traits that can be studied in isolation or in various combinations so the number of possible experiments is in the millions. The traits are all represented graphically so the student can observe the phenotypes directly. For instance, if the student selects the white eye mutation for the female parent their picture of the female parent will change to have white eyes. After the mating they will get a picture of the different progeny, along with numbers beside each picture to indicate the number of progeny of that type (number of females with white eyes, females with red eyes, etc.) The program includes a Chi Square calculator for doing statistical tests of the students hypotheses, and a notebook for recording results, observations, hypotheses and conclusions. Students can import the numerical results from their crosses and statistical tests directly into the notebook.

Using this program students can discover or study most of the important principles of Mendelian genetics, including dominant and recessive alleles, sex-linkage, lethal alleles, independent assortment, epistasis, linkage, gene order, linkage groups, and linkage maps. More importantly, students can discover these principles by doing the same sort of experiments as the original researchers, only much faster. The program is appropriate for a wide range of biology courses as the assignment determines the level of difficulty. Students can do statistical tests, but this is not required. Students can do complicated crosses with multiple traits, or simple crosses with only one trait at a time. If a student is confused by a complicated cross, they can always do some additional simpler crosses to try to figure out what is going on. They can also do additional crosses with the progeny from their crosses, and their progeny, etc. This ability to devise their own experiments and try many different permutations is a major strength of the FlyLab.

PedigreeLab

PedigreeLab will generate numerous pedigrees for a particular genetic disease. The student can examine the pedigrees to determine the inheritance pattern of the particular disease. In addition, the student will be able to examine various molecular markers and determine whether they are linked to the genetic disease. This is a key process in the current search for human genetic disease genes and is normally very difficult to explain to students. Having them actually go through the process should significantly improve their learning of these difficult concepts.

TranslateIT

So far, the Biolabs project has produced two molecular biology simulations. The first, TranslateIT, http://www.cdl.edu/TranslateIT, simulates some of the original experiments used to crack the genetic code, one of the key discoveries in molecular biology. These experiments rely on radioactive materials and difficult to produce RNA templates so they can't be done in the normal biology lab. Students design and create simple RNA molecules in the simulation that they then translate in a virtual in vitro translation mix. The program shows a simple animation of the techniques that would be used to analyze the products of the translation and then gives them the amino acid sequence of any proteins produced in their experiment. The student must logically analyze the results of multiple experiments to deduce the properties of the genetic code, just as the original researchers did, only with the advantage of being able to do experiments in minutes that normally take months to carry out. Various properties of the code that can be determined using this simulation are the triplet nature of the code, that the code is non-overlapping, codon assignments for particular amino acids, and the existence and identity of stop codons.

HemoglobinLab

In the second molecular biology simulation, the HemoglobinLab, http://www.cdl.edu/HemoglobinLab, students investigate various aspects of the molecular biology of hemoglobin, using case studies. The goal is for the student to learn how changes in the nucleotide sequence of a gene may effect the protein sequence, which may effect the structure of the protein, which may effect the function of the protein, which may effect the properties of the cell, which may in turn effect the physiology of the individual. Students choose a case by selecting a patient from a pull down menu with a list of over a dozen patients. For each case the students can examine the doctors notes about the symptoms and medical history of the patient, examine the color of a vial of the patients blood, examine the blood under a microscope to see if there are changes in the red blood cells, run a sample of the blood on an electrophoresis gel to determine if there are physical changes in the globin protein, and, finally, the student can determine the amino acid sequence of the patients globin protein. Having determined the sequence of the protein the student can go to the DNA sequence editor and try to alter the DNA sequence of the normal gene to see what type of DNA mutation would cause the changes found in the patient they are examining. The patients have a variety of mutations in the globin gene ranging from simple point mutations that change one amino acid, such as in sickle cell anemia, to deletions and insertions causing frameshifts, such as some of the thallasemias. The mutations cause many different patient phenotypes, such as anemia, brown blood, polycythemia (too many red blood cells), and purple skin color.

DemographyLab

The DemographyLab, http://caldera.calstatela.edu/DemoLab, models human population growth in several different countries around the world. Students can use this lab to investigate how differences in population size, age-structure, and age-specific fertility and mortality rates affect human population growth. Default values for seven countries have been incorporated into the program to allow comparisons between nations with very different demographics, such as Japan and Nigeria. In addition, students can change any of the parameters to create their own experiments. The proportion of males and females in each five year age group, the total population size, the mortality rate for males and females in each five year age group and the birth rate per female in each five year age group can all be set by the student using a simple graphical interface. After running the simulation for 100, 200 or 300 years, students get summary statistics for the population at the end of the time interval, such as life expectancy, birth rate, population growth rate, etc. They can view a line graph of population numbers over the course of their experiment, see a graphical representation of the population structure for every five years of the experiment, or examine the number of males and females in each age group for each five year period. Using the program a variety of demographic phenomena can be demonstrated, such as exponential growth and decline, stable age structure, zero population growth, demographic momentum, dependency ratios, sex ratios and marriage squeezes.

MitochondriaLab

Three biochemistry and cell biology labs currently in development are MitochondriaLab, GlycolysisLab, and PlantLab. MitochondriaLab will simulate electron transport, proton gradients and oxidative phosphorylation in mitochondria. Students will be able to recreate the classic experiments that established the chemiosmotic theory as the mechanism for energy production in the cell. They will add various substrates and inhibitors to their virtual mitochondrial extracts and then measure the consumption of oxygen over time. From their results they can work out some of the steps in the pathway and the mechanism by which the chemical energy is converted into ATP molecules.

GlycolysisLab

In GlycolysisLab students will study the enzymatic reactions of the glycolysis pathway and the kinetics of biochemical reactions by varying quantities of substrates and cofactors and then measuring reaction rates.

PlantLab

The PlantLab will simulate the photosynthetic reactions in leaves. Students will vary wavelength and intensity of light, CO2 and oxygen concentrations, temperature, etc. and then measure the consumption of CO2 and the production of sugar in their simulated leaves.

ReproductionLab

The tenth and last lab currently under development is ReproductionLab. This lab will simulate the human female reproductive cycle. Students will be able to inject various hormones, alter diet and exercise, and remove organs in their virtual subjects and then can track hormonal changes and various physiological responses over time. Student experiments should lead to a greater understanding of hormone action, positive and negative feedback loops, homeostasis, reproductive cycles and physiological pathways. Understanding this physiological cycle is obviously of great importance to the students and will also teach them many important physiological concepts.

Assessment

The precursor to all of these labs is the original virtual Fly Lab. This simulation of fruit fly genetics is now used in biology classes all over the world, and has created so much demand on the server hosting the program that there are now five different mirror servers. The Virtual Fly Lab, EvolveIT and TranslateIT have been field tested in an upper division genetics course with encouraging results. 98% of the students in this course considered their Virtual Fly assignments useful in learning genetics, 83% found EvolveIt to be useful and 93% found TranslateIt useful. Some sample comments from the students are:

"Excellent 3 part demonstration of Mendelian genetics - each lesson built on the previous one - excellent practice for the tests. Also includes repetition of important concepts (X linkage, etc.)"

"Liked best - actually enjoyed! (what a concept) Virtual Fly I, II, and III. These were great fun to puzzle out - someone's going to hate me for saying this"

"TranslateIt was enjoyable because it requires the student to investigate and solve the problem."

"Virtual Fly and TranslateIt were the assignments I got the most out of. I liked the way it made you systematically think to solve the problems."

"I liked the Virtual fly and EvolveIt activities because they allowed you to do some investigation on your own and they made you think about what was really happening, which made you understand the material better."

There were only a few negative comments, usually having to do with the difficulty of getting on-line and using the programs. Students who are uncomfortable with computers are at a disadvantage when using these simulations and special care must be taken to make sure they get the most out of the simulations. The only other negative comment was, "I liked TranslateIt the least because it made my head hurt." While this is unfortunate, if the BioLabs project can produce more simulations that cause some students heads to hurt, then the project will be producing simulations that change, for the better, the way biology is taught.

Acknowledgements: The following individuals contributed to the development of these simulations, their contributions are gratefully acknowledged: Bob Desharnais, Steve Wolf, Zed Mason, Ron Quinn, Terry Frey, David Hanes, Judith Kandel, Nancy Smith, Sally Veregge, Abbe Barker, Michelle LaMar, Chuck Schneebeck, Rachel Smith, Lou Zweier, Scott Anderson, Peilin Nee and Anne Scanlan-Rohrer. Partial support was provided by U.S. National Science Foundation grant DUE 9455428 to Bob Desharnais.

This document is maintained by: Jeffrey R. Bell
Last Update: Thursday, October 29, 1998