A sample notebook write-up appears further down.
In this module, you’ll be examining all the assumptions and happenstances that allowed Gregor Mendel to deduce his simple rules of inheritance. Briefly put, all 7 pea loci he examined displayed simple dominance, were unlinked and on autosomes (the non-sex, nuclear chromosomes), and did not have any interactions with each other or weird functions. In this module, we’ll examine some of the fascinating violations to these rules (as Mendel did, much to his chagrin, when he attempted to show the generality of his findings in a rather more bizarre plant).

How it will run:
You will do

3 x 10 pointers (0 or 1 non-Mendelian phenom)
2 x 15 pointers (0, 1 or 2 non-Mendelian phenoms)
1 x 20 pointer (either multiple genes or multiple alleles) Do not proceed on this until you have heard the full explanation in class.

- There are no groups per se, but you are welcome to consult with each other to your heart’s content.
A sample write-up is included at the end of this presentation. Note that it relies upon the chi-square test (introduced in a tutorial in the module).
- For each problem, you will need to identify a homozygous dominant and homozygous recessive organism for each gene you invoke (except, of course, in the cases where it would be dead...). If you have any sense at all, these will arise naturally in the course of your investigations anyway.
- You do not need to chi-square or report #s on crosses showing 3:1 segregation, but you’ll lose points if you MISS something and haven’t reported your findings there.
- As always, the software has many time-saving features, but the most important one is sitting on your shoulders. Don’t be seduced by the lure of brain-dead clicking and sorting of butterflies. Have a goal. Formulate a testable hypothesis, design a clean, clear test, and proceed in an orderly fashion. Also, while it’s cool to do a cross with as many things segregating in it is possible, it’s probably not your best bet for finding something out. Focus. Focus focus. Focus focus focus...

The scenarios:
Nada:
Everything as Mendel observed and explained. 'Course, you have to argue that this is the case, not just assert it. Besides being right, you must demonstrate that your 3 loci are not demonstrating linkage.

Linkage
Purpose: Investigating the ties that bind.
Relevance: Each gene doesn't get its own chromosome. Contrariwise, each one has neighbors. If the genes you happen to be studying are close enough, they will no longer behave independently--where one goes, the other will tend to follow. That tendency, of course, is directly proportional to how close the genes are. The closer they are, the more unlikely a recombination will separate the two, and the more likely it is that the parental configuration will be represented in the gametes.
Starting Point: Several simple phenotypes, two of which will display linkage, which should be detectable by a distortion of the ratios Mendel saw. If acting independently, two genes showing simple dominance should segregate 9:3:3:1 in a cross between heterozygotes: 9 being the dominant/dominant phenotypes, the two 3s representing the two dominant/recessive phenotypes, and the 1 being organisms expressing both recessive traits.

Codominance
Purpose: Seeing in shades of gray.
Relevance: The phenotypes examined by Mendel were all binary--his pea plants were either tall or short, white or red flowered, etc. That's not always the way it goes--sometimes heterozygotes have intermediate phenotypes of the two alleles that make them up.
Starting Point: One locus has a pair of codominant alleles; giving rise to non-3:1 segregation, and more than two phenotypes.

SexLinkage
Purpose: Girls get to have all the fun.
Relevance: In organisms where sex is determined by unlike chromosomes, gene distributions can be unequal. In humans, guys get ripped off because the Y chromosome is largely a desert; functionally this renders all genes on their single X as homozygous. This is why some genetic difficulties, such as red-green colorblindness, are more common in males.
Starting Point: Butterflies with one locus on the X chromosome. Note that in actuality, butterfly sex determination is the OPPOSITE of ours, with females getting the short end of the chromosome (XY). HOWEVER, I have chosen to model them like us: XX females, XY males, in order to keep things familiar.

Mitochondrial Linkage
Purpose: Dads are deadbeats. Not only do they make a minimal investment prior to birth, genetically speaking they also shortchange their offspring. The mitochondrial genome (yes, Virginia, there is one, though it doesn't have a vast number of genes on it) is solely the mother's contribution.
Relevance: It happens, and it has been exploited to trace maternal lines in trying to deduce the origin and migration of humans (a la 'mitochondrial eve'. Russia's last Czar, Nicholas, was also ID'd (rather, his remains) based on a rare exception to the 'mother only' inheritance situation.
Starting Point: You guessed it--one of the genes in play is on the mitochondrial genome.
Homozygous dominant lethality
Purpose: Recognize the presence of a lurking lethal allele.
Relevance: How would you know if all the butterflies of dominant phenotype were heterozygotes?
Starting Point: The usual--several strains of butterflies, and one locus where inheritance of both dominant alleles is lethal. No dead butterflies show up--you have to 'see' what's missing by considering what you observe with what you'd observe if all phenotypic classes were represented.

Combined Lethality
Purpose: More studies on genetic interactions.
Relevance: Revealing more of the simplifications involved in Mendel's assumptions. Sometimes there are cases where an organism can happily tolerate defects in one pathway or another, but not both. In such cases, certain expected combinations of traits simply will not be found.
Starting Point: Some strains with a lurking problem: an inviable potential situation which you must discover--by NOT finding it!

Homozygous Dominant Lethality
Bad things sometimes come in twos. In some cases, such as sickle cell anemia, being homozygous is not immediately lethal, though life expectancy is severely shortened. In others, no progeny with two defective/deleterious copies of a gene are observed. Having a missing phenotypic class can be detected because it skews the lovely ratios predicted by simple Mendelian genetics. While in most cases, like sickle cell, the recessive allele is the one that is homozygous lethal, in the case modeled here it's the dominant one (this is because recessive lethals in these butterflies are 'invisible' embryos that never see the light of day; there would be no way to distinguish a recessive lethal trait from a purebreeding population of homozygotes).

Maternal Effects
Purpose: Delve deeper into the mysteries of the origin of phenotype!
Relevance: As the saying goes, it ain't who you are, it's who you come from. The mother provides much of the early environment for an embryo. Indeed, many of the early processes in the embryo are mediated by material that is wholly maternal in origin and has nothing to do with the genotype of the offspring. In such cases, any phenotypic outcomes reflect the MATERNAL genome, not that of the father or child.
Starting Point: A set of loci, one of whose effects is MATERNALLY determined. Which is it, and how can you find it?

Segregation Distortion
Purpose: Natural selection is based on one extremely straightforward underlying force: the more kids you leave behind, the more prevalent the components of your genome will be in the future. What, then, would happen if an allele 'figured out' how to 'win' by killing off or outcompeting other alleles even before an organism had been made?
Relevance: Such things happen, and one of the forms is called 'segregation distortion'. The mechanism is through the destruction of sperm bearing one allele of a given gene by those bearing a different allele. Obviously, the allele doing the destroying gets off to a quick head start in the Darwinian game of natural selection! There are other, less lethal versions of this strategy out there; the topic of sperm competition is a hot one in biology right now, and rightly so--in the race to the egg, there is only one winner, and it doesn't matter if sperm #2 would've produced Arnold Schwarzenegge

Example write up
This is somewhat dated but still has merit; if I don't get a chance to re-write it, listen to the intro lecture!!

Please note two key changes since this example was written:
1) There is now a uniform nomenclature for genes/phenotypes. You can access it in 3 ways. First, when you hold down the <SHIFT> key and move the mouse over any butterfly, it show you its phenotype. The 3 letters you should use to refer to that phenotype (and the genotype underlying it) are capitalized and bold. Thus Overall Wing Color should be referred to as OWC. Second, when using the Evaluator room or the GenoPeek tool (only available when not logged in) the genotypes are reported in this way. Finally, there is a listing of genotypes under the Protocols, Menus... menu.

2) While the notebook will NOT require an overall purpose and rationale, each segment should be presented in that fashion. To facilitate this, note that there is a button in the notebook labeled 'New Section'. You should use this whenever you start a new argument, such as when you set out to show that you have sex linkage, etc. Thus the notebook will be a group of structured arguments about the phenomena you found or disproved.

A good write up for your work here should include the following key components:
--An overview, stating what, if any, non-Mendelian patterns you observed and for what traits you observed them.
--a presentation section, in which you briefly detail key findings and crosses that gave rise to these observations. Conclusions drawn should be justified and accompanied by chi-square analysis or bullet-proof reasoning (for example: I conclude that x and y phenotype are lethal in combination because in examining 100 progeny of the type Aa bb x aa Bb, no aa bb offspring were observed)
--an appendix where you give the location of strains homozygous for genes mentioned. Part of the purpose of this is to force you to make and recognize these strains, which is virtually an inevitable part of your work anyway. You should also identify whether the allele in play is recessive or dominant. Note these butterflies do not individually need to be homozygous for EVERYTHING; only for a given trait.

EXAMPLE:
In this species, I determined that top wing size is inherited as a codominant trait. Bottom wing vein color and wing color both display simple dominance, but they are autosomally linked.
Purpose: Testing for codominance of Top Wing Size
In studying top wing size, I created strain 101 (homozygous for small wings, the dominant trait TWS/TWS) and 102 (homozygous for large wings, the recessive trait tws/tws). Strain 103 and 104 are heterozygous for the trait, having arisen from the 101 x 102 cross (TWS/tws).
Rationale: For a hypothesis of codominance, I predict that the heterozygotes will have a phenotype distinct from the two parents, and a cross of two heterozygotes will show a 1:2:1 phenotypic ratio corresponding to the 1:2:1 genotypic ration (1 TWS/TWS : 2TWS/tws : 1tws/tws)
Observations:
I crossed strains 103 x 104 and observed
24 large wing (tws/tws)
49 medium wing (TWS/tws)
27 small wing (TWS/TWS)
By chi-square analysis, given expectations of 25:50:25, I derive
1^2/25 + 1^2/50 + 2^2/25 = .22
The chi-square value for 2 degrees of freedom for p = .1 is 4.6, so my results are well within the expected variance range, i.e. my hypothesis accounts for the my observations well. Thus I conclude that the trait is codominant.
In studying bottom wing vein color (BVC) I found it to segregate 3:1, as did overall wing color (OWC). I created strain 110 which is homozygous for blue veins, the recessive trait and for red wings, the recessive (hence owc/owc,bvc/bvc) and strain 112, homozygous for Green veins and Blue wings (OWC/OWC,BVC/BVC). Crossing these two yielded strains 113 and 114, male and female of genotype (OWC/owc,BVC/bvc).
The cross between 113 and 114 is predicted to yield a 9:3:3:1 ratio of offspring as follows:
9 Green Wing, Blue vein (OWC/?,BVC/?)
3 Green Wing, green vein (OWC/?,bvc/bvc)
3 red wing, Blue vein (owc/owc,BVC/?)
1 red wing, green vein (owc/owc,bvc/bvc)
However, in a cross yielding 112 offspring, I observed
66 (OWC/?,BVC/?)
12 (OWC/?,bvc/bvc)
12 (owc/owc,bvc/bvc)
24 (owc/owc,bvc/bvc)
The large number of the last class suggests that there is linkage involved. Chi-square analysis demonstrates that 9:3:3:1 segregation is almost certainly NOT occurring, in that the chi-square value is
3^2/63 + 2^2/7 + 4^2/7 + 17^2/1 = 292
the chi-square value for 3 degrees of freedom for p < .01 is 11, thus my findings will appear by bad luck (way, WAY) less than one time in 100, and linkage is a reasonable hypothesis.
APPENDIX:
strain 101 is homozygous for small wings (TWS)
102 is homozygous for large wings (tws)
112 is homozygous for Green veins (BVC) and Blue wings (OWC)
114 is homozygous for blue veins (bvc) and red wings (owc)

Some free advice:
--Try the link from the main Mendelstar page to the NDSU site. It’s nice.
--Take small bites and chew your food before you swallow--while creativity and bold thinking play a critical role in science, they don’t call it a discipline for nothing. If you don’t learn to organize your thoughts and make them work for you, they’ll never be any good to anybody. A brilliant hypothesis is a place to begin; from that point, you should be your own harshest critic.
--Think first, think twice, do once. As with previous scenarios, you can spend an infinite amount of time spinning your wheels. In some ways, that’s the whole point. Get a good grasp of what you might be seeing and how you intend to recognize it before you go snark hunting!
--Dealing with the software: It may be the age of the portable supercomputer, but not in the BLC. You’re smarter than those computers, so cut them some slack and keep your clicking to a dull staccato.
Also: some of you WILL discover series of actions that I never in my wildest dreams would’ve imagined anyone would ever undertake. The software won’t be ready for these. You can deal with this best in 2 ways:
PREVENTION: Don’t do bizarre stuff. If you don’t have any idea what to do, don’t do anything and certainly don't click everything. The tools have intended uses and they CAN work at least when applied in ways I have anticipated!
REPAIR: If something goes wrong, be a scientist about it. What were your most recent actions, or what were you doing differently than you had done before? What were the surrounding circumstances or settings? Make careful observations: what was the exact error message you received, what did the program do as it expired? The more information you can give me, the more likely I will be able to reproduce the problem, and anything that can be reproduced is subject to scientific methods--and can be fixed.

Inhibition
Purpose: Dissecting a pathway with inhibitors.
Relevance: Much of any cell's time and enery is spent on regulation, and it's not all positive. Pathways are commonly turned OFF from a default ON state. What does this show up as genetically?
Starting Point: One of the three pathways in play has a gene product acting as an inhibitor. To simplify things, you can make these assumptions:
1) The inhibitor always completely shuts off the pathway if a wild type copy is present
2) The inhibitor will manifest as a simple dominant
Two Step Pathways
Purpose: Mendel got away with one--be the Super Mendel!
Starting Point: Several genes, one of which harbors a killer allele. Which is it?
Relevance: In choosing characters that looked vastly different, Mendel did himself a huge favor--he selected traits that were the result of single genes acting in unrelated pathways. In modern science, folks are generally interested in the exact opposite case--they wish to study the components of a single pathway intensively. Of course, this means that the simple 3:1 segregation patterns of yore won't hold up.
Starting Point: One of your pathways has two steps. To further add sparkle to your day, you may or may not be able to see a unique phenotype for the intermediate step, i.e. if your pathway looks like:
(A) => (B) => (C)
then you MAY not be able to observe a unique phenotype if the pathway is missing its second step, i.e. is stuck at B. It may appear identical to a pathway stuck at A. In such cases, you'll have to rely on other means to spot the two-stepper!
Complicated Two step pathways
Purpose: Pulling out all the stops (steps?)
Relevance: Allowing you to act out your masochistic tendencies without inflicting bodily harm.
Starting Point: One of your pathways gets to be as complicated as it wants to be. It will consist of two steps; that you can be sure of. But the intermediate product may or may not manifest a unique phenotype (if it doesn't, it will look just like the starting product), there may or may not be an inhibitor for either step, and the genes performing the steps may or may not be varying (i.e. the gene for step one, for example, may always be 'on', though its inhibitor may be represented in your library as wild type and mutant forms). Have fun!