21 topics
Genetics help, one cross at a time
A genetics unit goes wrong at the moment the 4-square Punnett grid becomes a 16-square one, because the grid was never the point. The point is that alleles at unlinked loci assort independently, which means you can multiply single-trait probabilities instead of drawing anything. Students who learned the grid as a picture stall here. Students who learned it as probability keep going into linkage, pedigrees and population genetics without a wobble.
Where students get stuck
The 16-square Punnett grid for a dihybrid cross takes me forever
Stop drawing it. For unlinked genes, treat each trait as its own coin flip and multiply. A cross of AaBb by AaBb gives 3/4 A_ and 3/4 B_, so 3/4 times 3/4 is 9/16 showing both dominant traits, and 1/4 times 1/4 is 1/16 showing both recessive. That is where 9:3:3:1 actually comes from. The same trick scales: a trihybrid would need 64 squares, but three multiplied fractions take about fifteen seconds. Only draw a grid when the genes are linked and independent assortment no longer holds.
My chi-square value is huge and I do not know what I did wrong
Almost always one of two things. First, chi-square must be run on raw counts, never on percentages or ratios. Plugging in 75 and 25 as percentages when you actually counted 300 and 100 flies changes the statistic completely, because the whole test is sensitive to sample size. Second, degrees of freedom is the number of phenotype classes minus one, not the number of offspring minus one. A monohybrid cross has two classes and one degree of freedom, so your critical value at p equals 0.05 is 3.84. A large chi-square is not automatically an error, either. It is evidence that independent assortment is not happening, which in a linkage question is the answer.
Incomplete dominance, codominance and epistasis all look the same to me
Look at what a heterozygote shows, then at whether one gene is masking another. Incomplete dominance blends: a red and a white snapdragon give pink, and the F2 ratio is 1:2:1 because the heterozygote is now visibly its own class. Codominance shows both alleles fully and separately, which is why an AB blood type carries both A and B antigens rather than something in between. Epistasis is different in kind, because it involves two genes, not two alleles. One locus shuts down the expression of another, and the giveaway is a modified 9:3:3:1, such as 9:3:4 or 9:7. When a ratio has fewer classes than you expected, suspect epistasis.
In a pedigree I can never decide if the trait is recessive or X-linked
Work through three questions in a fixed order. Does the trait skip a generation, with two unaffected parents having an affected child? Then it is recessive, because both parents carry it silently. Is it recessive and strongly skewed towards males? Then suspect X-linked, since a male needs only the one copy he has. Now find the killer test case: an affected father with an unaffected daughter rules X-linked recessive out entirely, because he would have passed his only X to her. For dominant traits, every affected child must have an affected parent, so a single affected child of two clear parents rules dominance out on its own.
Recombination frequency and map units do not click
One map unit is one percent recombinant offspring, and that is the whole definition. The work is deciding which offspring are recombinant. Set up a testcross against a fully recessive parent so every gamete from the parent of interest shows up directly in the phenotype. The two largest offspring classes are parental, meaning no crossover happened, and the two smallest are the recombinants. Divide recombinants by total offspring and multiply by 100. With three genes, the two rarest classes are double crossovers, and comparing them to the parental classes tells you which gene sits in the middle. Anything at or above 50 percent means the genes assort independently, so distance can no longer be measured this way.
What's covered
Genetics topics you can work through with a tutor, generate practice on, or turn into flashcards and a study plan.
Mendelian inheritance
- Monohybrid and dihybrid crosses
- Law of segregation and independent assortment
- Test crosses and back crosses
- Probability rules and the forked-line method
Beyond simple dominance
- Incomplete dominance and codominance
- Multiple alleles and ABO blood groups
- Epistasis and modified phenotype ratios
- Polygenic traits and continuous variation
- Penetrance and expressivity
Chromosomes and linkage
- Meiosis, crossing over and gamete variety
- Sex-linked inheritance and pedigrees
- Recombination frequency and gene mapping
- Nondisjunction and karyotype disorders
Molecular genetics
- DNA replication, transcription and translation
- Mutation types and reading-frame shifts
- Operons and gene regulation
- PCR, gel electrophoresis and CRISPR basics
Population genetics
- Hardy-Weinberg equilibrium and its assumptions
- Allele and genotype frequency calculations
- Selection, drift and gene flow
- Chi-square goodness of fit tests
Genetics questions
Can it check a Punnett square or a pedigree I have already drawn?
Yes. Hold your page up to the camera or share your screen, and it reads what you have drawn. On a pedigree it will follow your reasoning generation by generation and tell you where the deduction broke, which is usually one unaffected parent who has to be a carrier. On a Punnett square it checks your gametes first, since a wrong gamete row makes every cell downstream wrong.
Will it just give me the answers to my genetics problem set?
No, and it will push back if you ask. On a cross it will ask you what the parental gametes are before it fills anything in, because that is the step marks are actually lost on. If you are out of time and stuck, ask for one worked example of the same problem type, then do yours beside it. You can also generate a practice set on the same topic that comes back with full worked solutions.
Does it cover the maths side, like chi-square and Hardy-Weinberg?
Yes, and it treats them as maths rather than as recall. For Hardy-Weinberg it starts from the fact that you are usually given q squared, the recessive phenotype frequency, and have to take a square root before anything else works. It will also state the five assumptions the equilibrium rests on, since a question asking why a population is not in equilibrium is really asking which assumption broke.
I am in a first-year university genetics course, not high school. Is it useful?
Yes. The harder material is where explaining out loud helps most: three-point mapping, complementation tests, chi-square on linkage data, and the difference between a nonsense and a missense mutation at the level of the codon table. You can work a mapping problem on the whiteboard with it and check the gene order against the double crossover classes before you commit to an answer.
Stuck on genetics right now?
Talk it through out loud, share your screen, and watch it worked out step by step on a whiteboard.
Start free — no card