how close is close enough?

131
How close is close enough? Part II Mendel vs 1000 Ideal Worlds Build the world in BioBIKE biobike.csbc.vcu.ed u This demonstration is best viewed as a slide show, enabling you to simulate a session and make changes in cursor position more obvious. To do this, click Slide Show on the top tool bar, then View show.

Upload: torie

Post on 19-Jan-2016

49 views

Category:

Documents


1 download

DESCRIPTION

How close is close enough?. Part II Mendel vs 1000 Ideal Worlds. Build the world in BioBIKE biobike.csbc.vcu.edu. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: How close is close enough?

How close is close enough?Part II

Mendel vs 1000 Ideal Worlds

Build the world in BioBIKEbiobike.csbc.vcu.edu

This demonstration is best viewed as a slide show,enabling you to simulate a session and make

changes in cursor position more obvious.To do this, click Slide Show on the top tool bar, then View show.

Page 2: How close is close enough?

Define an ideal genetic world

pp p

male-genotype female-genotype

Pp Pp

- Define genotypeHow?

Page 3: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

How?

Page 4: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

- Define joining of gametes to form progeny

How?

Page 5: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

- Define joining of gametes to form progeny- Define how color is determined

How?

Page 6: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

- Define joining of gametes to form progeny- Define how color is determined

- Define Mendel's experiment 929

crosses

Page 7: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

- Define joining of gametes to form progeny

- Define how color is determined

- Define Mendel's experiment

Go to the BioBIKE Portalbiobike.csbc.vcu.edu

Page 8: How close is close enough?

Go to a public site

Page 9: How close is close enough?

Your name (no spaces)

Enter a log in nameClick New Login

Page 10: How close is close enough?

The BioBIKE environment is divided into three areas as shown. You'll bring functions down from the function palette to the workspace, execute them, and note the results in the results window

Function palette

Workspace

Results window

Page 11: How close is close enough?

Two very important buttons on the function palette:

On-line help (general)

Something went wrong? Tell us!

HELP!

PROBLEM

Page 12: How close is close enough?

Two very important buttons in the workspace:

Undo (return to workspace before last action)

Redo (Get back the workspace you undid)

Page 13: How close is close enough?

pp p

male-genotype female-genotype

What now?

The first step in making the ideal world is to define what are male- and female-genotypes.

Mouse over the DEFINITION button

Page 14: How close is close enough?

Click DEFINE to bring down the DEFINE function.

Page 15: How close is close enough?

A DEFINE function box is now in the workspace.

Before continuing with the problem, let's consider what function boxes mean.

Page 16: How close is close enough?

General Syntax of BioBIKE

Function-name Argument(object)

Keyword object Flag

The basic unit of BioBIKE is the function box. It consists of the name of a function, perhaps one or more required arguments, and optional keywords and flags.

A function may be thought of as a black box: you feed it information, it produces a product.

Page 17: How close is close enough?

• Function-name (e.g. SEQUENCE-OF or LENGTH-OF)

• Argument: Required, acted on by function

• Keyword clause: Optional, more information

General Syntax of BioBIKE

• Flag: Optional, more (yes/no) information

Function-name Argument(object)

Keyword object Flag

Function boxes contain the following elements:

Page 18: How close is close enough?

General Syntax of BioBIKE

Function-name Argument(object)

Keyword object Flag

… and icons to help you work with functions:

• Option icon: Brings up a menu of keywords and flags

• Clear/Delete icon: Removes information you entered or removes box entirely

• Action icon: Brings up a menu enabling you to execute a function, copy and paste, information, get help, etc

Page 19: How close is close enough?

General Syntax of BioBIKE

Function-name Argument(object)

Keyword object Flag

… and icons to help you work with functions:

• Option icon: Brings up a menu of keywords and flags

• Clear/Delete icon: Removes information you entered or removes box entirely

• Action icon: Brings up a menu enabling you to execute a function, copy and paste, information, get help, etc

And now back to our show…

Page 20: How close is close enough?

The DEFINE function asks for two things: the name of the variable to be defined and the value it is to be given. Let's call the variable male-genotype. The value of male-genotype will be "Pp"

Click on the variable argument box to open it up for entry…

Page 21: How close is close enough?

… and type male-genotype (remembering to connect the words with a hyphen), closing the box afterwards by pressing Tab.

Page 22: How close is close enough?

Tab closes the entry box and automatically opens the next one (if it exists).

Type the genotype "Pp" (be sure to include the quotation marks) and press Enter.

Page 23: How close is close enough?

The definition is now complete. But it will not take effect until the function is executed

Click the Action icon.

Page 24: How close is close enough?

… and click Execute.

Page 25: How close is close enough?

Notice that a new VARIABLES button appears. We'll use it later to access the newly defined variable.

Notice also that the definition is confirmed by showing it in the Results Window

Page 26: How close is close enough?

Repeat what you did, this time defining female-genotype in the same way

Page 27: How close is close enough?

Now to define meiosis.

Mouse over the DEFINITION button and click DEFINE, just as you did before.

Page 28: How close is close enough?

Click the name box…

Page 29: How close is close enough?

…type in meiosis, and press Enter.

Page 30: How close is close enough?

If you like, you can give a short summary in English what your function does, that is, what is meiosis.

To do this, mouse over the summary icon…

Page 31: How close is close enough?

… click summary, and enter the summary of the function in the text box that appears. Finally, press Enter to accept the summary.

Page 32: How close is close enough?

How to define meiosis?

It requires the diploid genotype of the organism and then chooses one of the chromosomes to go into a haploid gamete.

The argument (input) to the function is the genotype. Click the argument box…

Page 33: How close is close enough?

… and enter the word genotype, a symbol that will represent whatever genotype the function is given. Then press Enter.

Page 34: How close is close enough?

What does meiosis do?

You'll define this by putting the appropriate action into the form box. Click that box….

Page 35: How close is close enough?

…how to choose a chromosome at random? BioBIKE has a CHOOSE-FROM function that will serve our needs.

Mouse over the ALL button…

Page 36: How close is close enough?

The ALL button provides BioBIKE functions in alphabetical order.

Mouse over the C menu item…

Page 37: How close is close enough?

…and click CHOOSE-FROM.

Page 38: How close is close enough?

CHOOSE-FROM what?

From the genotype!

Click the sequence box…

Page 39: How close is close enough?

… and type genotype and press Enter.

Page 40: How close is close enough?

The definition of meiosis is now complete. To make it active, mouse over the Action Icon…

Page 41: How close is close enough?

…and click Execute.

Page 42: How close is close enough?

Notice that a new FUNCTIONS button appears. We'll use it in a moment to access the newly defined function.

Notice also that the definition is confirmed by showing it in the Results Window

Page 43: How close is close enough?

Try out your new function.

Mouse over the FUNCTION button and click Meiosis.

Page 44: How close is close enough?

Click on the genotype box of Meiosis.

Provide it with male-genotype, by mousing over the VARIABLES button and clicking male-genotype.

Page 45: How close is close enough?

Execute your function as you do any other function, by mousing over the Action Icon…

Page 46: How close is close enough?

…and clicking Execute.

Page 47: How close is close enough?

The result of meiosis now appears in the Results Window.

Execute the function a few more times to get a feel for what it does.

Page 48: How close is close enough?

With Meiosis in hand, you can now define the male and female gametes.

Bring down a DEFINE box as you have previously and set the variable as male-gamete. (Be sure to remember the hyphen – you can't put spaces within variable names).

Then click the value box.

Page 49: How close is close enough?

We want to define male-gamete as the product of meiosis working on male-genotype.

You can bring down a Meiosis box into the value box, but it's easier to cut/paste the box you already have.

Mouse over the Action Icon of the Meiosis function in the Workspace.

Page 50: How close is close enough?

…and click cut.

Page 51: How close is close enough?

Then mouse over the Action Icon of the value box…

Page 52: How close is close enough?

…and click paste.

Page 53: How close is close enough?

Now you can execute the definition.

To define female-gamete, click on the word male-gamete, and when the box is open for editing (turns white), change male to female. Press Enter to accept this change.

Make the same change to male-genotype and execute the function.

Page 54: How close is close enough?

Now to define the joining of the male and female gametes to form a progeny.

Bring down a DEFINE box, and type progeny as the name of the variable.

Click on the value box. We need a function that will join the gametes. Fortunately, BioBIKE has such a function. Find it using the ALL button.

Page 55: How close is close enough?

To enter the male-gamete and female-gamete into the two arguments of JOIN, click each argument box and find the variable on the VARIABLES menu.

Page 56: How close is close enough?

Once the function is complete, execute it several times.

Why is the behavior different with this function than you saw when you executed Meiosis several times?

Page 57: How close is close enough?

Now we need to define a function that analyzes a genotype and decides what color it determines.

Bring down a DEFINE-FUNCTION box as before and fill in the holes.

Page 58: How close is close enough?

But how to make the decision?

IF the genotype is pp then the flower is white. Otherwise the flower is purple.

BioBIKE makes such decisions using the IF-TRUE function, available from the FLOW-LOGIC menu.

Page 59: How close is close enough?

IF-TRUE looks at a condition (in this case whether the genotype is the same as "pp") and THEN does something if the condition is true ELSE it does something else.

Let's first consider the condition. Click that box…

Page 60: How close is close enough?

…and bring down the SAME function, available from the FLOW-LOGIC button, Logical-Comparison submenu.

Page 61: How close is close enough?

We want to know if the genotype is the same as "pp".

Click on x, type genotype, and press Tab. Then type "pp" into the y box.

Page 62: How close is close enough?

BioBIKE will ordinarily ignore the distinction between upper and lower case. So the difference between "pp" and "Pp" will be ignored…

…unless we tell BioBIKE to consider case.

Mouse over the Option Icon…

Page 63: How close is close enough?

…and click CASE-SENSITIVE

Page 64: How close is close enough?

If the genotype indeed is the same as "pp", THEN the color should be "white". Type "white" in the THEN box.

ELSE the color should be "purple". Type "purple" in the ELSE box.

(Be sure to press Tab or Enter after each entry)

Finally, execute the completed DEFINE-FUNCTION.

Page 65: How close is close enough?

Bring down a COLOR-OF box from the FUNCTIONS menu, and execute the function a few times with different genotypes (within quotation marks). Are the results as you expect?

Page 66: How close is close enough?

Now to teach BioBIKE how to do a cross.Bring down a new DEFINE-FUNCTION.Call the function Make-Flower.It will return a flower from a Pp x Pp cross.

It won't need any required arguments, but how many functions in the body will be required to do the job? Let's think…

Page 67: How close is close enough?

We need to define male-gamete and female-gamete. That's two functions.

We need to combine them into the progeny.

We need to determine the color of the progeny.

That's four. So we need four holes. Mouse over the Options Icon of the Body Form…

Page 68: How close is close enough?

…and click Add two more forms.

Then mouse over the Options Icon again and click Add another to get a total of four forms to fill in.

Page 69: How close is close enough?

Click on the first form and DEFINE male-gamete as before (or cut and paste the existing box).

Then put DEFINE female-gamete in the second form.

Then DEFINE progeny in the third.

Then put COLOR-OF progeny in the fourth.

Page 70: How close is close enough?

Execute DEFINE-FUNCTION .

Then bring it down the new MAKE-FLOWER function from the FUNCTIONS button and try it out a few times.

When you're satisfied, clear the screen (x in upper right corner).

Page 71: How close is close enough?

929crosses

MAKE-FLOWER is good for one cross, but Mendel did 929. We need to repeat the MAKE-FLOWER function 929 times.

You can repeat the function that many times by going to the DEFINITION button…

Page 72: How close is close enough?

… and clicking REPEAT-FUNCTION.

Page 73: How close is close enough?

Click the function box and fill it with MAKE-FLOWER… not the words but the function. To get the function…

Page 74: How close is close enough?

… go to the FUNCTIONS button.

Page 75: How close is close enough?

Then click the number box and fill it with 929.

Page 76: How close is close enough?

Execute the function to do Mendel’s complete experiment in this ideal genetic world.

Page 77: How close is close enough?

And there's the results, all 929 of them. How many of the flowers are white?

Counting 929 virtual flowers is not something humans are good at, so let the computer do the work…

Page 78: How close is close enough?

…Bring down the COUNT-OF function from the ALL button.

Page 79: How close is close enough?

Click the query argument box, and fill it with the word "white" (including the double quotation marks).

Page 80: How close is close enough?

Count of "white" in what? We need to specify what should be counted.

To do that mouse over the Options Icon…

Page 81: How close is close enough?

…and click the IN option.

Page 82: How close is close enough?

Again,… Count of "white" in what? In this case, it's the results, i.e. that list of "purple" and "white" flowers.

In general, however, we're looking for the count of "white" in the results of Mendel's experiment.

To bring Mendel’s experiment (i.e. the REPEAT-FUNCTION box) into the value box, you could cut/paste as before, but this time, try drag/drop. Click under REPEAT-FUNCTION for a half second and hold the button down…

Page 83: How close is close enough?

Now drag the box over to the value box…

Page 84: How close is close enough?

…so that the upper left corner of the dragged box is in the target box (the target box will gain a red outline), then release the mouse button.

Page 85: How close is close enough?

Now the function is set up to tell you how many flowers in Mendel’s experiment are white.

Page 86: How close is close enough?

Execute it. In fact, execute it twice.

Page 87: How close is close enough?

You may get a different answer each time, because if Mendel had repeated his experiment with 929 flowers, he would probably have gotten a different number of white flowers. Each time is potentially different.

Page 88: How close is close enough?

That’s Mendel’s experiment.

Let’s package it as a function to make it easier to manipulate.

Page 89: How close is close enough?

Call the function MENDELS-EXPERIMENT (no apostrophe’s please!), and give it a summary that will remind you what it does.

Page 90: How close is close enough?

Mendel’s experiment is precisely the COUNT-OF box we just made.

To bring that box into MENDELS-EXPERIMENT, drag it into the form box of the Body section…

Page 91: How close is close enough?

…and drop it there.

Page 92: How close is close enough?

Now execute the function definition to add it to your collection of functions.

After doing that, you might want to clear the screen (red X in upper right corner).

Then Bring down MENDELS-EXPERIMENT from your FUNCTION button and execute it a couple of times.

Page 93: How close is close enough?

Each time you execute the function, you will probably get a different number of white flowers counted. That’s the way experiments work.

Page 94: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

- Define joining of gametes to form progeny- Define how color is determined

- Define Mendel's experiment 929

crosses

We’ve now accomplished everything we’ve set out to do,… except one thing: We still don’t know whether Mendel was right to call his results 3:1.

Page 95: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

- Define joining of gametes to form progeny- Define how color is determined

- Define Mendel's experiment

Recall… he observed 224 white flowers.

A 3:1 ratio predicts 232¼ white flowers.

Is that close enough?

With our virtual experiment in hand, we’re finally ready to define what we mean by “close enough”.

224 705 + = 929

232¼ 696¾

Observed:

Expected: + = 929

Purple White Total

Page 96: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

- Define joining of gametes to form progeny- Define how color is determined

- Define Mendel's experiment

Suppose that Mendel is right. The ratio of purple to white flowers is really 3:1, and his deviant result was just because experiments give different results each time, and 224 is the result he happened to get. If he did it again, he might get 232 white flowers.

Well, we now can do it again! If we repeat the experiment many times, and his observed value of 224 is common (as in the case below), then Mendel was justified in making his approximation.

Purple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

Page 97: How close is close enough?

Define an ideal genetic world

- Define genotype- Define meiosis

- Define gamete

- Define joining of gametes to form progeny- Define how color is determined

- Define Mendel's experiment

On the other hand, suppose Mendel is wrong. In that case his observed value (or worse) will very rarely crop up (as in the other case below).

We have a strategy! Just repeat his experiment and see how frequently we get 224 white flowers (or fewer).

Purple flowersN

umbe

r of

exp

erim

ents

Replications in 3:1 world

Observed

Expected

Purple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

Page 98: How close is close enough?

We already have the experiment.

All we have to do is to repeat it – say 100 times.

You know how to repeat a function. Go to the DEFINITION button and get REPEAT-FUNCTION.

Page 99: How close is close enough?

Now drag MENDELS-EXPERIMENT into the function box…

Page 100: How close is close enough?

…and release it.

Page 101: How close is close enough?

We want to repeat the experiment a lot of times, but too many times and execution will be painfully slow.

Insert 100 as the number of times.

Then execute the function

Page 102: How close is close enough?

Well there’s our answer.

All we need to do is to count how many times the number of white flowers is 224 or less.

But humans weren’t made to count so many numbers. We need to filter away the numbers we don’t want and keep the ones we do.

Page 103: How close is close enough?

To do this, bring down the FILTER function…

Page 104: How close is close enough?

…and drag (or copy/paste) the result at the bottom of the screen into the data box….

Page 105: How close is close enough?

…and drop it there.

Page 106: How close is close enough?

We want the filter to discriminate between those numbers less than or equal to 224 and the rest, so the test we want is <= (less than or equal to).

Mouse over the choose test icon…

Page 107: How close is close enough?

…and click <= (less than or equal to).

Page 108: How close is close enough?

Now type 224 into the value box.

Page 109: How close is close enough?

…and execute the function.

Page 110: How close is close enough?

That’s a whole lot better.

But still I’d rather not count (estimate, yes; count, no).

Fortunately, I have a COUNT-OF function available. All I need to do is go to the Action Icon of FILTER…

Page 111: How close is close enough?

…and click Surround-with.

Page 112: How close is close enough?

Now I bring down COUNT-OF, as I have done in the past…

Page 113: How close is close enough?

…and FILTER is surrounded by the COUNT-OF function.

Execute that to get the answer of…

Page 114: How close is close enough?

22. 22 times out of a 100 repetitions of his experiment, Mendel would be expected to get a count of white flowers at least as deviant from expectation as he actually observed.

Almost. 224 is 8¼ flowers deviant from the expected 232¼ ,… but so is 240½!

We should also consider as equally deviant those experiments that give more than 240 white flowers.

Page 115: How close is close enough?

We could construct from scratch a different filter and count what passes through it, but it’s much easier to copy and modify the current function.

So copy it…

Page 116: How close is close enough?

…and go to the FILE button to paste the copy into the workspace.

Page 117: How close is close enough?

Now, alter the test so that it is…

Page 118: How close is close enough?

Now, > (for greater than 240)

Page 119: How close is close enough?

…and change the threshold value from 224 to 240.

Page 120: How close is close enough?

Execute…

Page 121: How close is close enough?

So 22 times out of a hundred Mendel would have gotten a number of white flowers that is too low, at least as deviant as the number he actually got.

And 24 times out of a hundred he would have gotten a number that is too high, at least as deviant as the number he actually got.

22 + 24 = 46% of the time… almost half the time. Not so unlikely!

Page 122: How close is close enough?

So Mendel WAS right in calling his result 3:1. Or at least his observed value was not out of line with the expected value, given normal variation.

Purple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

22%24%

Page 123: How close is close enough?

In considering such problems as this one (VERY common!), you generally don’t have access to a computer that can do this simulation.

But you do have access to something that’s equivalent…

Purple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

22%24%

Page 124: How close is close enough?

Statistics!

The chi-squared (X2) test does through an equation exactly what you just did on the computer.

Try it!

Chi-Squared Test

Where Oi is the observed value

And Ei is the expected valuePurple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

22%24%

Page 125: How close is close enough?

Application of Chi-squared test

(705 – 696¾)2 (224 – 232¼)2

696 ¾ 232¼

= 0.391

Chi-Squared Test

Where Oi is the observed value

And Ei is the expected value

+X2=

Purple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

22%24%

224 705 + = 929

232¼ 696¾

Observed:

Expected: + = 929

Purple White Total

Page 126: How close is close enough?

Application of Chi-squared test 0.391 Chi-Squared Test

Where Oi is the observed value

And Ei is the expected value

X2=

Purple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

22%24%

Now go to a web site that will translate this number into an area under a curve as in the figure to the left.

http://www.fourmilab.ch/rpkp/experiments/analysis/chiCalc.html

Page 127: How close is close enough?

Scroll down to the calculator…

Page 128: How close is close enough?

Insert .391 for the X2 value

…and 1 for the degrees of freedom (the flower is either white or it is not… one question to answer)

…then click Calculate.

Page 129: How close is close enough?

Chi-Squared Test

Where Oi is the observed value

And Ei is the expected valuePurple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

22%24%

The probability (area under the curve) you get from the X2 test should be within several percentage points from what you got from repeating the virtual experiment…

…because that’s what a X2 test is! (so long as you do the experiment an infinite number of times!)

Page 130: How close is close enough?

Chi-Squared Test

Where Oi is the observed value

And Ei is the expected valuePurple flowers

Num

ber

of e

xper

imen

ts

Replications in 3:1 world

ObservedExpected

22%24%

Page 131: How close is close enough?