| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
SPECIAL TOPICS |
* Animal Science Department, University of California, Davis 95616;
Division of Animal Physiology, The University of Nottingham, Loughborough, Leics, UK
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONTYPICAL RESULTS AND DISCUSSIONDISCUSSIONLITERATURE CITED |
|---|
Key Words: computer simulation model • classroom • dairy cow • metabolism • nutrition
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONTYPICAL RESULTS AND DISCUSSIONDISCUSSIONLITERATURE CITED |
|---|
Introduction to Systems Theory
A system is a regularly interacting or interdependent group of items forming a unified whole. Systems theory entails identifying interacting elements within a defined system and exploring relationships among those elements and how they function together to change system behavior (Wiener, 1948
). The systems approach is particularly useful in teaching nutrition because understanding of nutrient digestion, metabolism, and use cannot be considered separately. For instance, an animal responds constantly to the changing nutrient supply with which it is provided and can only grow if it has adequate energy and substrate supply for protein synthesis. Physiological processes such as growth, lactation, and reproduction respond to nutrient supply and, in turn, affect nutrients available for other physiological processes. If an animal is the system and nutrients are elements of the system, nutrients must be identified and interactions among nutrients represented explicitly in order to understand and predict the animal’s response to dietary changes. A representation of nutrient flow in an animal is presented in Figure 1. Nutrients are ingested, digested, and metabolized in support of growth, reproduction, lactation, heat production, etc., and resulting waste products are excreted. Changes in nutrient flows for one process change nutrient availabilities for other processes and overall response of the animal. Nutrient reservoirs or pools are represented as the quantities of nutrients available for physiological processes (units of mole, gram, etc.), and the flows or fluxes of nutrients into and out of these pools are represented by arrows. Nutrient flux rates have units of quantity per unit time (mole/day, grams/minute, etc.).
|
Molly is a dynamic, mechanistic model of digestion and metabolism of a lactating dairy cow. The model was described in detail by Baldwin (1995) and in earlier publications (Baldwin et al., 1987a,b,c). Based on initial BW, fat percent (or body condition score), length of simulation (days), nutrient analysis of diet(s) fed, and a parameter describing a cow’s genetic potential to produce milk, the model estimates changes in BW, weights of viscera and body fat, milk and milk component production, waste excretions, respiratory exchange, energy costs of individual nutrient transactions, ration metabolizable energy values, and total heat production.
The digestion element of Molly was originally developed to provide estimates of nutrient digestion for the animal metabolism element and has been discussed in several previous publications (Offner and Sauvant, 2004; Johnson et al., 2005; Hanigan et al., 2006; Johnson and Baldwin, 2007). This paper is focused on use of the animal element of Molly as an aid in teaching animal metabolism. Thus, discussion of the digestion element is not included here beyond mentioning a few basic features. The animal element of the model (Figure 2) begins with absorbed nutrients and defines transactions associated with 10 state variables. Total amino acids (TAa) are made up of 4 groups of amino acids: His, Lys, sulfur amino acids (Met and Cys), and all other amino acids (Aa). Other whole body nutrient pools are glucose (Gl), acetate (Ac), and fatty acids (Fa). Body composition state variables are body protein (Pb), visceral protein (Pv) and storage triacylglycerol (Ts). Milk output state variables are milk protein (Pm), milk lactose (Lm), and milk fat (Tm). Inputs to nutrient pools are estimates of rates of absorption of the nutrients generated by the digestion element or defined as model inputs based on experimental data and rates of formation from other nutrients or body stores. Outputs from nutrient pools described are oxidation, conversions to other nutrients, and use for the synthesis of body proteins and fat or secreted products (milk, milk fat, etc.).
|
, the incorporation of total amino acids into body protein in body (suffix B) is TAaPbB and the conversion of glucose to milk lactose (Lm) in the udder, a component of viscera (suffix V) in the model is GlLmV. Inputs and outputs for a pool represent metabolic transactions computed in the animal element of the model. The main nutrient transaction equations, units, and definitions shown graphically in Figure 2 are presented in Table 1. Fluxes are calculated using kinetic, usually Michaelis-Menten type, equations justified and parameterized on the basis of experimental data collected using isolated cells, tissue slices, and arteriovenous difference techniques. Example data used are presented in introductory lectures and included in Table 1.
|
). This allows SAa, Lys, His, or Aa to limit the synthesis of milk (Pm), body (Pb), and visceral (Pv) proteins and, based on rates of 
-lactalbumin synthesis, lactose synthesis. Stoichiometric parameters, which define the products of amino acid degradation in the model, are dynamic variables dependent on amounts of individual amino acids entering and leaving the several pools. Sources of amino acids are digestion of microbial protein, rumen bypass and abomasally infused proteins and amino acids, and degradation of body and visceral proteins. Individual amino acids leave the pools for the synthesis of milk, body, visceral, salivary, fetal, and placental proteins, and via amino acid degradation. Stoichiometric parameters are calculated based upon metabolic pathways for degradation of individual amino acids. Embedded in flux equations for metabolic transactions are explicit use of current understanding of nutrient metabolism and hormonal effects on metabolism. For example in Equation 2, Table 1, the term FaTsF represents fatty acid incorporation into storage triglyceride in adipose and is calculated using a Michaelis-Menten type equation. Maximal velocity of fatty acid incorporation into triglyceride (VFaTsF) is scaled to empty BW to the 0.75 power (EBW0.75) and is modulated by an affinity constant for fatty acid uptake by adipose (KFaTsF) relative to plasma fatty acid (cFa) and for fatty acid uptake relative to anabolic hormone (Ahor), injected insulin (INS), and plasma glucose (cGl).
 |
Students manipulate model parameters and run simulations through a user-friendly program. Model parameters are set in the simulation input data window (Figure 3). Parameters that can be changed include length of simulation (Go until day XXX), milking frequency, diet, feed intake level, recombinant bovine somatotropin (BST) injections, ionophore, day of conception, amino acid and casein infusions, milk production potential (udder cells), initial BW, and body fat percent or initial body condition score. Simulations can easily be paused at any stage of lactation to change diets or feed intake level and then continued. Included in the program are feed ingredient and complete mixed diet libraries and tools for viewing model outputs. Feeds are added to the feed library using embedded regression equations that convert nutrient analyses such as CP, ADF, NDF, triglyceride, lipid, urea, lignin, and ash to nutrient inputs needed by Molly (Johnson et al., 1998). Diets can be created and saved in a user-defined diet library using the feed library. Model outputs for a single time point can be viewed from a list of predefined diagrams or for model changes over time using a graphing function. Continuous numerical data outputs can also be saved for additional data plotting and analyses.
|
, the process of writing state equations describing metabolism and physiology of the cow is illustrated. Several example equations are examined in detail to illustrate how knowledge gained in biochemistry courses is applied to living systems. Example discussions of equations used to describe metabolic processes are presented in the following section.
|
The reductionist method focuses on testing a hypothesis and understanding function through reducing it to its smallest components. Similarly, the systems approach also focuses on testing a hypothesis but achieves understanding of function by building on lower level knowledge to predict higher level function. For example, if a goal was to predict body protein changes in a dairy cow during lactation, the systems approach would start at describing the supply of limiting amino acids for body protein synthesis. Figure 4 is a depiction of sulfur amino acid metabolism as represented in Molly. Sources of sulfur amino acids are from the diet and rumen micobes (absSAa), from protein degradation in the body (PbSAaB) and viscera (PvSAaV). Uses of sulfur amino acids are body protein synthesis (SAaPbB), visceral protein synthesis (SAaPvV), milk protein synthesis (SAaPmV), salivary protein synthesis (SAaSAL), degradation to supply carbon for carbohydrate or fatty acids (SAaDEG), and to support pregnancy (SAaPRG). Each of these terms is interdependent. For instance, a decrease in sulfur amino acids from the diet (rumen) could result in an increase in protein degradation, a decrease in protein synthesis, a decrease in sulfur amino acid degradation, or a combination of these to compensate for the decrease in sulfur amino acid supply. Interdependence occurs because the equations representing these transactions are dependent on the concentration of sulfur amino acids in blood, and in turn, blood sulfur amino acid is dependent on the supply and uses of sulfur amino acids.
The lowest level of function represented in Molly is equations representing the kinetics of sulfur amino acid entry and use (absSAa, PbSAaB, SAaPbB, etc.). The next level is the interaction of processes involving sulfur amino acids as shown in Figure 4. Figure 5 shows how sulfur amino acid processes interact with other nutrients and contribute to overall metabolism of the dairy cow. Based upon the transactions represented in Figure 4, an equation is written describing the input and output of sulfur amino acids for metabolic processes (equation 7 in Table 1).
|
). Similar equation sets are used to represent lysine, histidine, and all other amino acids in equations 5, 6, and 8 in Table 1, respectively. A further discussion of how to interpret limiting amino acids predictions is included in sample simulation 3. Sample Simulations Illustrating Metabolic Concepts
Student projects begin with identification of a particular problem or question, developing a hypothesis to be evaluated, running simulations to challenge the hypothesis, and analyzing the simulation results to determine whether the hypothesis is tenable. Three sample problems/questions are posed to students to help familiarize them with use of the model. The first example posed is an analysis of BST effects upon metabolism and production, the second is evaluation of effects of diet changes and amino acid supplementation on N efficiency, and the third is the effect of supplementing with limiting amino acids. These are presented herein to illustrate the use of Molly to teach a systems approach to nutrition and metabolism. Table 2 contains initial simulation settings that are entered in the input simulation window (Figure 3) for each example. Additional settings in the input simulation window specific to each example are described in each section. Nutrient (Table 3) and amino acid compositions (Table 4) for each diet used in the examples can be accessed from the standard diet and user-defined diet lists that come with the Molly program. The Molly installation program, printouts of all simulations results (in bmp format), and class handouts used in the following examples are available at http://animalscience.ucdavis.edu/research/Molly
|
|
|
In class, students are first introduced to using the Molly simulation program by running a set of 8 simulations comparing the effects of 2 different diets and 2 potential milk production levels with and without BST administration (based on Chalupa et al., 1996). Student reports follow the format presented below.
BST Objectives (Stated in Class Handout and Repeated in Report)
Demonstrate milk production responses to BST administration to cows of average and high production potential fed 2 diets according to the same feeding strategy. Evaluate profitability of BST in these feeding and production situations and evaluate how BST affects metabolism.
BST Methods
Run simulations at 2 different milk production potentials (1,000 and 1,500 udder cells), with 2 different diets (Diet1 and Diet2), with and without BST for a total of 8 simulations. Table 5 contains settings for each simulation that should be used in the input simulation window (Figure 3) in addition to the settings listed in Table 2. Students collect data from model output at d 84 and 308 of lactation. Therefore simulations must be paused at d 84 (see "Go Until Day" column, Table 5) and then continued to d 308 by changing the "Go Until Day" value to 308 in the input simulation window and using the "run" command under the model menu at the top of the program (see "Model Command" column, Table 5). A BST value of 1.0 represents a basal level of BST. To simulate BST administration, this value is reset to 3.0, representing 3 times basal. For simulations with BST, simulations 5 to 8, BST is given at 70 d of lactation. Therefore, to start a simulation, "Go until Day" is set to 70, and under the model menu, reset and then run is pressed. At 70 d, BST is changed to 3.0, and "Go Until Day" is changed to 84. Under the model menu, run is pressed to continue the simulation. At 84 d, data are collected, "Go Until Day" is changed to 308, and under the model menu, run is pressed to finish the simulation.
|
|
TYPICAL RESULTS AND DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONTYPICAL RESULTS AND DISCUSSIONDISCUSSIONLITERATURE CITED |
|---|
). Diet2 resulted in higher milk production (300 to 400 kg) and higher milk protein (12 to 26 kg) compared with Diet1. Diet2 is higher in starch (2%) and insoluble protein (2%), which contribute to increased lactose production (GlLmV), increasing milk production and milk protein production. Overall, higher producing cows (1,500 udder cells) respond less to BST in milk and milk protein production than lower producing cows (1,000 udder cells).
|
Assuming the cost of BST is $0.40/d and given the estimated profit over feed cost estimated by Molly over a 308 simulation, BST increased profit by approximately $500 per 308 d of lactation. Diet1 was more profitable than Diet2 by $50 per lactation, and increasing genetic merit from 1,000 to 1,500 increased profit by approximately $800 per lactation.
2. Nitrogen Efficiency Simulation Exercise
Short simulations beginning at d 80 and ending at d 94 of lactation in which dietary crude protein level, starch, and digestibility of rumen insoluble crude protein are adjusted are used to demonstrate variation in nitrogen utilization in terms of milk production and nitrogen excretion (based on Marini and Van Amburgh, 2005).
Nitrogen Efficiency Objectives
Identify which diet (level of rumen undegradable protein and crude protein) is optimum in terms of milk yield, nitrogen excretion, and rumen function. Explore how changes in crude protein, and rumen protein degradability affect rumen function, total nitrogen excretion, and partitioning of nitrogen among urine, feces, and milk.
Nitrogen Efficiency Methods
Students run a total of 40 short simulations including 2 diets (Diet3 and Diet26) at 4 levels of crude protein (16, 18, 20, and 22%) and 5 levels of rumen insoluble protein digestibility (dpif) for a total of 40 simulations. Input simulation window (Figure 3) settings are shown in Table 7 in addition to initial simulations settings for this example (Table 2). When "short" is selected as the length of simulation, input simulation settings will automatically change to the appropriate initial BW, body fat, and "Go until day" values for a short simulation. Time values will display as days 1 to 14 even though the simulation is actually simulating d 80 through 94 of lactation. Udder cell number is 1,500, and data are recorded only at the end for all simulations. Therefore, "Reset to time=0" and then "Run Simulation" are selected from the model menu to run each simulation.
|
must be entered in the "Enter ACSL CMD" box in the input simulation window (Figure 3). For instance, to run simulation 5, the phrase "s dpif=1.8, fdpi=0.061, fdst=0.254" should be entered in the box. Dietary starch is "fdst" and insoluble protein is "fdpi" both in kg/kg of diet. Adjusting insoluble protein changes total crude protein in the diet and because the diet nutrient composition must sum to 1, starch must be adjusted inversely to insoluble protein. The insoluble protein digestibility factor, dpif in kg/d, alters the rate of insoluble protein hydrolysis to amino acids in the rumen. Therefore, increasing dpif decreases rumen undegradable protein (RUP). Nitrogen Efficiency Results and Discussion
For Diet3 simulations, responses due to increasing crude protein and RUP are consistent (rows 1 to 20, Table 8). Increasing RUP within crude protein level increases milk, milk protein, nitrogen in milk (NMilk), and nitrogen in feces (NFeces). Nitrogen in urine (NUrine) is highest at 22% crude protein, 0.237 RUP. Cellulose digestion and microbial population are highest at 22% crude protein, 0.268 RUP, indicating that microbial production is maximized at this level. Although milk production and milk protein production are highest at 22% crude protein, 0.447 kg/kg of RUP, the most milk, milk protein, and nitrogen in milk per crude protein % is produced with 16% crude protein, 0.279 kg/kg of RUP.
|
). Diet26 is higher in nonfiber carbohydrates and in cellulose and hemicellulose than diet3. Milk production follows a similar pattern to microbial production. This connection shows how Molly represents the reality that higher amounts of fermentable carbohydrate in the diet maintain microbial fermentation. Thus, this type of diet increases milk production more than does supplying a higher amount of crude protein and RUP. Milk production and milk protein production are highest at 20% crude protein, 0.329 kg/kg of RUP. Nitrogen in milk and rumen microbes is highest at 22% crude protein, 0.313 kg/kg of RUP, but cellulose digestion is highest at 22% crude protein, 0.367 kg/kg of RUP. Efficiency of nitrogen use is highest for milk and milk protein per crude protein % with 16% CP, 0.252 kg/kg of RUP. Microbial production is positively correlated with dietary crude protein level. However, within crude protein level, microbial production is maximized when RUP composes one-third or less of total dietary crude protein. In general, increased microbial synthesis corresponds with higher digestion coefficients for structural carbohydrates and an increased need for nitrogen for microbial growth. Increased microbial production increases milk, milk protein, nitrogen in milk slightly, nitrogen in feces, and nitrogen in urine if RUP levels are moderate. Therefore, diet composition affects nitrogen efficiency relative to crude protein and RUP level, and an increase in microbial production does not always correspond with an increase in nitrogen efficiency.
Within a crude protein level, increasing RUP changes the excretion of nitrogen from urine to feces and increases nitrogen in milk. Therefore, if nitrogen efficiency is based on optimizing milk profit, the best combination of RUP and crude protein will be the one that maximizes milk output over the cost of handling nitrogen output. If costs of nitrogen in urine and feces are ignored, nitrogen is used most efficiently for milk and milk protein production on diet3 16% crude protein, 0.279 kg/kg of RUP, which corresponds with a greater amount of nitrogen in milk and a lower (but not the lowest) amount of nitrogen excreted in urine and feces relative to crude protein intake.
3. Limiting Amino Acid Supplementation Exercise
Nine diets with different amino acid compositions are simulated for different lengths of time to discover which amino acids are limiting for milk and milk protein production. Diets are supplemented with the appropriate amino acid and simulated again for a total of 18 simulations. Diets HLDRUP, HHDRUP, LHDRUP, and LHDRUPM are from Noftsger and St-Pierre (2003), and HPMU, LPLU, LPMU, LPHU, and LPHUUR are from Davidson et al. (2003).
Limiting Amino Acid Objectives
Quantify response of cows in simulations to diets differing only in amino acid composition. Explore how increasing amino acid supplementation changes milk, milk protein, body protein, and visceral protein synthesis, and demonstrate the limiting amino acid theory.
Limiting Amino Acid Methods
The lowest maximum potential rate of protein synthesis in each category is set as the most limiting for synthesis of that protein (Figure 5). Rates of use of all other amino acids for synthesis of that protein are calculated from this rate. Equations 5 to 8 (Table 1) are the calculation of actual protein synthesis for each process. Equations 9 to 12 (Table 1) show the selection of the lowest protein synthesis rate for each amino acid group as the rate of protein synthesis.
Figure 5 lists the maximum potential protein synthesis rates from a 2-wk simulation starting at 80 d in milk. From the simulation results, sulfur amino acids are most limiting for body protein synthesis, milk protein synthesis, and -lactalbumin synthesis, whereas Lys is most limiting for visceral protein synthesis. Different amino acid groups may be limiting at different times during lactation. However, supplementing with sulfur amino acids during this time period will increase body and milk protein synthesis and -lactalbumin synthesis until another amino acid group becomes limiting.
Simulation settings for the input simulation data window are listed in Table 2 (limiting amino acid rows only) and Table 9. Diet nutrient and amino acid content are listed in Table 3 and Table 4, respectively, and all diets are available from the user-defined diet list within the Molly simulation program. Abomasum infusions are set to the numbers indicated by entering the number only in the appropriate box next to the amino acid group in the input simulation data window (Figure 3). Once the "Go until day" values and udder cells have been set, diets have been selected, and abomasum infusions entered, simulations are run by selecting "Reset to time=0" and then "Run Simulation" from the model menu.
|
If sufficient limiting amino acid is added to a diet, the limiting amino acid changes to the next most limiting. For instance, in Table 10 simulation 9, diet HPMU is limited by sulfur amino acids for body protein, milk protein, and -lactalbumin synthesis (milk volume) and limited by lysine for visceral protein synthesis. Addition of an abomasum infusion of sulfur amino acids (simulation 10) changes the most limiting amino acid to lysine for milk protein and -lactalbumin synthesis and increases body protein, milk protein, and milk volume. Supplementing with a limiting amino acid increases the absorption of that amino acid (all simulations), increases milk nitrogen if the amino acid for milk protein was limiting (all simulations) and increases nitrogen in urine if amino acids are added in excess (i.e., casein infusions in simulations 1 to 2, 3 to 4, and 7 to 8 and excess sulfur amino acids in simulations 13 to 14 and 15 to 16). Nitrogen in urine decreases if an adequate amount of limiting amino acid is added because utilization of all amino acids for protein synthesis increases with addition of a limiting amino acid; therefore, amino acid degradation decreases (simulations 9 to 10, 11 to 12, and 17 to 18). Nitrogen in feces and microbial production remains approximately the same in all simulations unless microbial production increases; then fecal nitrogen output increases concomitantly.
|
-lactalbumin synthesis. Not enough casein is added to change the limiting amino acids for body protein or visceral protein synthesis. Visceral protein and body protein still increase because more amino acids have been supplied, but not enough to change their limiting amino acid. Similarly, in simulations 11 to 12, addition of sulfur amino acids is sufficient to change the limiting amino acid for milk and -lactalbumin synthesis to lysine, but not enough is added to change the limiting amino acid for body protein. Body protein still increases (because some limiting amino acid was supplied) and milk protein and milk volume increase, but visceral protein decreases because utilization of other amino acids were increased and lysine was not supplied. The increase in body protein, milk protein, and -lactalbumin synthesis decreased availability of other groups of amino acids, including lysine. Simulations 13 to 14 show results from supplementing with the limiting amino acid for visceral protein only. Visceral protein increases, but body protein, milk protein, and milk volume decrease or remain the same. Simulations 15 to 16 also show results from supplementing with the limiting amino acid for visceral protein (lysine), but visceral protein, milk protein, and milk volume increase because lysine is also very close to the most limiting for those processes. Simulations 17 to 18 show results from supplementing with 2 limiting amino acids, lysine, and sulfur amino acids. Supplementing with 2 amino acids increases utilization of all amino acids (urine nitrogen decreases and milk nitrogen increases) and milk protein and milk volume increases. But, body protein and visceral protein remain approximately the same. Therefore, adequate supplementation with a limiting amino acid for a specific organ such as the body, visceral, or mammary gland increases the protein synthesized for that organ. However, availability of nonlimiting amino acids for other processes will decrease because more total protein is being synthesized.
|
DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONTYPICAL RESULTS AND DISCUSSIONDISCUSSIONLITERATURE CITED |
|---|
Students are increasingly computer literate, with some universities and colleges now requiring computers for entering students. There is a dramatic increase in the use of teaching models in the biological and medical sciences as well. However, most students have little experience in the application of mathematics to biological systems, the concept of systems analysis, or how to develop models of biological function. Therefore, at least 1 laboratory session is used to familiarize students with the software including running the software, formulating diets, collecting results such as graphs, and discussing terminology. Usually a class project with the same set of inputs is most useful. In this situation, students can evaluate their individual output relative to that which is expected. This allows for troubleshooting and ultimately puts students on similar footing with an understanding of how to run the model and read the output. In addition, students are introduced to the help available within the software, and the teacher and teaching assistants hold office hours in the computer lab so students can drop by and get immediate help with running simulations. Using a computer model to teach nutritional concepts has the added advantages of allowing student learning to be self-paced and interactive.
Student Learning of Nutrition and Metabolism
Molly is a simulation model that allows students to explore impacts of feeding and management decisions, enabling them to integrate fundamental nutritional concepts. In so doing, students gain an appreciation of systems analysis, the quantitative nature of systems analysis, and develop an understanding of the power of simulation models when asking "what if" questions. In the present case, those questions relate to feeding strategies and the impact on cow production and metabolism. The specific learning goals of the examples described previously are to learn to critically evaluate profitability and metabolic impact of new technology (BST), to understand amino acid supplementation in reference to the limiting amino acid theory, and to understand how diets and processes impact N metabolism. Being able to interact with the model is beneficial to help the students understand these complex processes and gain an appreciation for the magnitude of metabolic changes that must occur when production or the nutritional regimen change. Prior to using the model, most students have had some intermediary metabolism in previous courses. However, the model can be used to describe production differences due to diet changes or to examine in-depth metabolic changes that must take place to change production as well as a quantitative description of the degree to which metabolism must change. The degree of detail used to examine simulation results can be dependent on the level of understanding of metabolism required for the course.
Teacher Management of Technology
Using a computer model to run experiments or ask "what if"-style questions can be challenging. Most students have had little exposure to applying mathematics to biological systems and will have difficulty at first understanding how substrates and products add up to represent changes in a metabolic pool (as described in the example with the sulfur amino acid pool). The role of the instructor in the managing this learning technology is important to the learning experience for the student. In using Molly in the classroom, the initial lecture(s) about the model should include an introduction to systems analysis and model development. Baldwin (1995) provides information required for such an introduction as well as more detailed descriptions of the modeling process. Students should be provided with a handout that the variables and the state variable equations (Table 1). Abbreviations used in the computer model can be confusing and add to students’ frustration over understanding metabolic concepts in a new format (Figure 4). However, a dictionary of terms is built into the model software, and tooltips are built into the software so that hovering over a term label produces a short explanation of the abbreviation. To aid in understanding the equations in Table 1 and basic functioning of the model, sample diagrams representing pools such as the sulfur amino acid pool example, described earlier, and numbers are given (Table 1) to show how processes add up and pool sizes change.
Gaining a complete understanding of the required simulation commands is also a challenge for some students. For example, determining when to use "Reset to time=0", "Run Simulation" and "Continue Simulation" commands in the model menu as used extensively in the BST simulations example. "Reset to time=0" sets the simulation back to parturition in a long simulation or to 80 d in milk in a short lactation and sets the parameters listed in the "initial simulation settings" in the input simulations window. But the only change that may be observed is the "Time (d)" box in the input simulation window will switch from a number to 0 and back to the original number. "Run Simulation" runs the simulation from the starting point (if reset was selected first) and "Continue Simulation" continues a simulation from the displayed time.
Other problems involve making sure that all simulation settings are correct before the simulation is run. The Molly program will pop up error messages if a diet has not been selected (by choosing a feed list and then pressing the "select a diet" button) or a feed intake scenario has not been chosen (from Intake 1 or Intake 2 drop-down lists, or both). Once students become comfortable with running simulations, collecting data, and presenting the results, using a computer simulation model offers an opportunity to learn how physiological function and nutrition affect cow performance.
Conclusions
Using a simulation model of metabolism can be a valuable way to teach metabolism. Benefits to the systems approach are students will learn how nutrient metabolism affects production, examine different feeding situations, think critically about nutrient, hormone, and metabolism and experience hypothesis development, data collection and analysis, and summarizing results. As with any new experimental technique or experimental approach, time will be needed to learn the software and become comfortable with the language used to describe Molly. But, the benefits will be that instead of memorizing metabolic pathways, students will learn how feeding nutrients, metabolism, and production fit together and apply what they have learned. These benefits can only be gained by using the systems biology approach.
1 Corresponding author: HAJohnson{at}UCDavis.edu
Received for publication July 3, 2007. Accepted for publication October 8, 2007.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONTYPICAL RESULTS AND DISCUSSIONDISCUSSIONLITERATURE CITED |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
