Biology by Bradford · IB Biology HL · D3.2.18–D3.2.21

When the ratio
comes out
wrong.

Genes on the same chromosome travel together. The offspring stop obeying Mendel: and that broken ratio turns out to be the most useful thing on the page.

01 · D3.2.18–D3.2.19 · The essentials

Three ideas,
then the maths.

Linked

Genes with loci on the same chromosome do not assort independently. They tend to be inherited as a unit, so a dihybrid cross behaves more like a monohybrid one.

Recombinants

Crossing over in prophase I can separate linked alleles. Offspring carrying combinations not present in the parents are recombinants: and they are always the minority classes.

Distance

The further apart two loci sit, the more likely a chiasma forms between them. Crossover frequency is a measure of distance: the basis of Morgan's first gene maps.

UNLINKED · DIFFERENT CHROMOSOMES LINKED · SAME CHROMOSOME A a B b assort independently in metaphase I A a B b travel together unless a crossover falls between them AB 25% Ab 25% aB 25% ab 25% AB 45% Ab 5% aB 5% ab 45% solid = parental outlined = recombinant DISTANCE SETS THE RECOMBINANT FREQUENCY P Q R P–Q close together: few recombinants P–R far apart: many recombinants
Same cross, two situations · unlinked genes give four gamete types in equal numbers, linked genes give mostly the two parental combinations
02 · D3.2.19 · Where recombinants come from

One crossover.
Four chromatids.

Green came from one parent, pink from the other. A chromatid that ends up two-toned is a recombinant. Tap through the parts.

PROPHASE I FOUR CHROMATIDS chiasma A a B b distance A B parental A b recombinant a B recombinant a b parental recombinants: the minority classes
One crossover

Two homologues, each already replicated into two sister chromatids. A single chiasma between the two loci affects only two of the four chromatids: which is why recombinants can never exceed 50% of the offspring.

03 · D3.2.20 · Linkage bench

Slide the loci
together.

A heterozygote AB/ab test crossed with ab/ab. Drag the loci closer and watch 1:1:1:1 collapse: and watch χ² decide, in real time, whether the difference is big enough to believe.

χ² = Σ (O − E)² E observed: the number you actually counted expected: predicted ratio × total offspring squaring removes the minus signs add the value up for every phenotype class dividing by E scales each gap against how big that class was supposed to be ONE PHENOTYPE CLASS AT A TIME O = 26 E = 30 O − E = −4 the gap χ² measures (−4)² = 16 square it 16 ÷ 30 divide by E = 0.53 this class’s term repeat for all four classes, then add the four terms together → χ²
Read it as a sentence: for every phenotype class, square the gap between observed and expected, divide by expected, then add them all up · the full five-step method follows in section 04
Start at 50%

At 50% the two loci behave as though they were on different chromosomes entirely: all four classes equal, χ² near zero, no evidence of linkage. Now start dragging.

04 · D3.2.21 · The chi-squared test

Five steps,
one number.

χ² asks a single question: is the gap between what you observed and what you expected too large to blame on chance?

Tap a step

The null hypothesis is always the boring one: there is no significant difference between the observed and expected frequencies, in other words, the genes are unlinked.

Run it
yourself.

Type your own observed counts. Everything else is calculated live: including the verdict.

Mendel's own data

The values loaded are Mendel's real counts from 556 F₂ pea seeds. Change any number and watch χ² respond: then try making the fit bad enough to cross 7.815.

AB Ab aB ab
Score the tray

Each token is one offspring from a test cross. A green box means the dominant allele of gene A, a filled dot means the dominant allele of gene B. Tally the four types, type your counts in, then check them: and only then find out whether these genes were linked.

df
1
2
3
4
5
critical value at p = 0.05
3.841
5.991
7.815
9.488
11.070

For every dihybrid cross in this course df = number of phenotype classes − 1 = 3

05 · Worked solutions

Two problems,
solved in full.

Do them on paper first, then open the solution. You can check every line in the calculator above.

Problem 1 · does it fit 9 : 3 : 3 : 1?

In tomato plants, purple stem (P) is dominant to green stem (p) and hairy stem (H) is dominant to smooth (h). Two plants heterozygous for both genes were crossed and 160 offspring were scored: 92 purple hairy, 26 purple smooth, 34 green hairy, 8 green smooth. Use a chi-squared test to determine whether these genes are linked.

Step 1 · hypotheses

H₀: there is no significant difference between the observed and expected frequencies, the genes are unlinked. H₁: there is a significant difference, the genes are linked.

Step 2 · expected values

Expected = predicted ratio × total. With 160 offspring: 9/16 × 160 = 90, 3/16 × 160 = 30, 3/16 × 160 = 30, 1/16 × 160 = 10.

Step 3 · the table
Phenotype
O
E
O − E
(O−E)²/E
purple hairy
92
90
+2
0.044
purple smooth
26
30
−4
0.533
green hairy
34
30
+4
0.533
green smooth
8
10
−2
0.400
Step 4 · sum and degrees of freedom

χ² = 0.044 + 0.533 + 0.533 + 0.400 = 1.51. There are four phenotype classes, so df = 4 − 1 = 3.

Step 5 · compare

At p = 0.05 with df = 3 the critical value is 7.815. Since 1.51 is well below it, p > 0.05 and the difference is not statistically significant.

Fail to reject H₀: the data are consistent with a 9:3:3:1 ratio, so there is no evidence that these genes are linked
Problem 2 · linkage and map distance

A fruit fly heterozygous for body colour and wing shape was test crossed with a black, vestigial-winged fly. The 300 offspring were: 142 grey normal, 12 grey vestigial, 15 black normal, 131 black vestigial. Determine whether the genes are linked, and if so, calculate the recombination frequency.

Step 1 · what was expected

A test cross on unlinked genes predicts 1 : 1 : 1 : 1, so each expected value is 300 ÷ 4 = 75. This is the prediction H₀ is built on.

Step 2 · the table
Phenotype
O
E
O − E
(O−E)²/E
grey normal
142
75
+67
59.85
grey vestigial
12
75
−63
52.92
black normal
15
75
−60
48.00
black vestigial
131
75
+56
41.81
Step 3 · sum and compare

χ² = 59.85 + 52.92 + 48.00 + 41.81 = 202.6, with df = 3. That is enormously greater than the critical value of 7.815, so p is far below 0.05.

Step 4 · identify the classes

The two large classes, grey normal and black vestigial, are the parental types, so the heterozygous parent carried the alleles in that arrangement. The two small classes are the recombinants: 12 + 15 = 27.

Step 5 · recombination frequency

RF = recombinants ÷ total × 100 = 27 ÷ 300 × 100 = 9%, which places the two loci roughly 9 map units apart on the same chromosome.

Reject H₀: the genes are linked, with a recombination frequency of 9%

A common slip: χ² is calculated from counts, never from percentages or ratios

06 · The bigger picture

Where this
fits.

Tap a concept to light up how it connects.

separated by produces counted as converted into detects Linked genes Crossing over Recombinants Recombinationfrequency Chi-squared Gene map
Overview

Linkage is invisible in a single organism. You can only see it in the numbers: which is why a statistical test sits at the centre of this topic rather than at the end of it.

Tap any concept to trace its connections · tap the background to reset

07 · Your turn

Fill in
the gaps.

Drag each term into the gap it belongs in. Two terms are traps.

Placed: 0 / 0
Genes whose loci are on the same chromosome aredrop
Linked alleles can only be separated bydrop
Offspring with combinations not seen in the parents aredrop
A test cross on unlinked genes is expected to givedrop
Expected value = predicted ratio ×drop
For any dihybrid cross the degrees of freedom isdrop
With df = 3 at p = 0.05 the critical value isdrop
If χ² is greater than the critical value, H₀ isdrop
08 · Check yourself

IB-style
multiple choice.

Single best answer, Paper 1 style. Have the critical value table handy.

Score: 0 / 0