Module 0: Vitis Evolution & Domestication

The cultivated grapevine (Vitis vinifera subsp. sativa) was domesticated ~8000 BP in the southern Caucasus and Transcaucasus (McGovern 2003) from the wild dioecious progenitor V. vinifera subsp. sylvestris. This module reviews the taxonomy of the genus Vitis, the Pinot Noir reference genome (Jaillon 2007), the dioecy→hermaphrodite transition that accompanied domestication, the 19th-century phylloxera crisis and its rescue by American rootstocks, and the MybA1 berry-color locus.

1. The Genus Vitis and V. vinifera

The genus Vitis contains roughly 60 interfertile species, distributed across temperate Eurasia, East Asia, and North America. Three geographic groups dominate viticultural and evolutionary discussion:

V. vinifera (Eurasian) — the source of essentially all premium wine grapes; ~10 000+ named cultivars worldwide.
North American clade — V. labrusca (concord, fox grapes), V. riparia, V. rupestris, V. berlandieri, V. cinerea, V. aestivalis. Phylloxera-tolerant roots; frost- and mildew-resistant.
Asian clade — V. amurensis (cold-hardy Manchurian grape), V. coignetiae, ~30 southern Chinese species.

The Pinot Noir Reference Genome

Jaillon et al. (2007) produced the first high-coverage grape assembly using a near-homozygous Pinot Noir line (PN40024, obtained by self-pollination). The genome spans 487 Mbp across 19 chromosomes (2n = 38), with ~30 434 predicted protein-coding genes. Key genomic features include a paleo-hexaploid ancestry shared with all rosids (three retained syntenic triplicates), an unusually rich complement of terpene-synthase genes underlying varietal aroma (linalool, geraniol in Muscat), and a large family of stilbene-synthase genes producing resveratrol as a phytoalexin.

\[\text{PN40024 reference: } 487 \text{ Mbp } \times 19 \text{ chromosomes } = 2n\!=\!38\]

~30,434 protein-coding genes (Jaillon 2007); paleo-hexaploid core eudicot ancestor.

Biogeography of the genus Vitis

2. Caucasus Domestication and the Dioecy→Hermaphrodite Shift

McGovern’s chemical analyses of Neolithic pottery sherds from Shulaveri-Gora and Gadachrili-Gora (Georgia) recovered tartaric-acid residues dated to ~8000 BP, the earliest chemically attested grape wine. Archaeo-genetic surveys (Ramos-Madrigal 2019, Dong 2023) place the primary V. vinifera domestication locus in the southern Caucasus, with a secondary, partly independent episode in the Levant/eastern Mediterranean producing the western “table grape” lineage.

From Dioecy to Self-Pollinating Hermaphrodite

Wild V. vinifera ssp. sylvestris is strictly dioecious—each vine is male or female and fruit-set requires cross-pollination. Cultivated ssp. sativa is overwhelmingly hermaphroditic: the flower simultaneously carries functional stamens and a receptive pistil, and the vine self-pollinates. This shift is controlled by a small (~140 kbp) sex-determination locus on chromosome 2. A female allele (F) and male allele (M) segregate in wild populations; hermaphroditism results from a recombinant H allele carrying the dominant stamen-producing function of M together with the functional pistil of F (Picq 2014, Massonnet 2020).

\[\text{Wild (sylvestris): } M/M, M/F, F/F \;\text{dioecious}\quad \Longrightarrow\quad \text{Cultivated (sativa): } H/H, H/F \;\text{hermaphrodite}\]

The transition was a major fitness innovation for the farmer: a single hermaphroditic vine reliably sets a full crop, and vegetative propagation (cuttings, layering) preserves genotype indefinitely. Myles et al. (2011) genotyped 950 unique V. viniferacultivars and found surprisingly limited pedigree depth—most elite varieties are sibs, parent-offspring, or grand-parents of a small ancestral clone pool.

The Domestication Bottleneck

Modern cultivar heterozygosity averages $H \approx 0.72$ whereas Caucasian wild sylvestris populations carry $H \approx 0.78$. The modest reduction in diversity reflects a long, mild bottleneck rather than a hard founder event—consistent with the coalescent model fit in Simulation 1, which places the effective bottleneck size around$N_b \sim 1000$. Vegetative propagation further limited drift by freezing rare clones into commerce for centuries.

3. The Phylloxera Crisis and the American Rootstock Rescue

In 1863 a mysterious decline of vines began in the southern Rhône valley. By 1890 nearly every vineyard in France, Italy, Spain, and Portugal had collapsed, with similar devastation from the Crimea to Australia. The proximate cause was the grape phylloxera (Daktulosphaira vitifoliae)—a root-feeding aphid native to eastern North America, inadvertently introduced to Europe via imported American vines.

The Root-Galling Biology

Phylloxera reproduces by cyclical parthenogenesis; the root form (radicicole) feeds by inserting stylets into cortical cells of young rootlets. In V. vinifera, feeding induces nodosities (young-root galls) and tuberosities (older-root galls) that become infected by secondary soil pathogens, killing the root system within 3–5 years. North American Vitis species evolved with the insect and deposit a necrotic periderm barrier that isolates and rejects the gall.

The Grafting Solution

The rescue was grafting: splice the V. vinifera fruiting scion onto a rootstock of American species (most commonly hybrids of V. berlandieri, V. riparia, and V. rupestris—e.g. Teleki 5C, SO4, Kober 5BB, 101-14 Mgt, 3309 Couderc). By 1910 essentially all European viticulture had been replanted on American rootstocks; today ~90% of all V. vinifera plantings globally are grafted. The cultivar (shoot system, fruit) remains pure vinifera, but the root system is an American species resistant to phylloxera.

\[\text{Graft} = \underbrace{V.\,vinifera\,\text{scion}}_{\text{flavor, chemistry}} \; \oplus \; \underbrace{\text{American rootstock}}_{\text{phylloxera resistance}}\]

4. Berry Color Genetics: The MybA1 Locus

Walker et al. (2007) mapped the dominant berry-color determinant in V. vinifera to a tight cluster of MYB transcription-factor genes on chromosome 2, notably VvMybA1. This MYB protein activates the anthocyanin biosynthesis pathway (UFGT in particular) in the berry skin. A single insertion of the Gret1 Ty3-like retrotransposon into the VvMybA1 promoter abolishes the gene’s expression. Homozygous insertion → no anthocyanin → white berries.

Single-Locus Mendelian Segregation

Because Ww heterozygotes are indistinguishable in phenotype from WWhomozygotes (colored), the ww (white) phenotype is fully recessive. An F2 cross of two heterozygotes therefore gives the classical 3 colored : 1 white ratio:

\[Ww \times Ww \longrightarrow \tfrac{1}{4}WW : \tfrac{1}{2}Ww : \tfrac{1}{4}ww \longrightarrow 3\,\text{Colored} : 1\,\text{White}\]

Pink, gris, and rouge berries arise from rare somatic reversions or partial revertants where one of the two ww copies partially regains expression in patches of the skin—most famously Pinot Gris, which is a bud-sport of Pinot Noir, and Pinot Blanc, a further bud-sport from Pinot Gris.

Muscat and Linalool Aroma

Complementary to MybA1, the Muscat-flavor locus encodes a point mutation in a 1-deoxy-D-xylulose-5-phosphate synthase (VvDXS1) that up-regulates monoterpene biosynthesis, yielding high concentrations of linalool, geraniol, and nerol. Muscat varieties are found in both colored (Muscat of Hamburg) and white (Muscat Blanc à Petits Grains) forms, confirming that MybA1 and Muscat segregate independently.

4b. Global Viticulture Today

Global vineyard area is ~7.3 million hectares (OIV 2022), down from a historic peak of ~10 Mha in the 1970s. The top producing countries by vineyard area are Spain (~0.95 Mha), France (~0.80 Mha), Italy (~0.70 Mha), China (~0.78 Mha of which mostly table grapes), Turkey (~0.40 Mha), and the United States (~0.40 Mha). Wine production is ~250 Mhl/yr, roughly two thirds of vineyard output; the remainder goes to table grapes, raisins, juice, and distillation.

The Latitude Band

Commercial vinifera viticulture historically occupies two narrow bands, 30–50°N and 30–45°S, defined by a growing-season average temperature of 13–21 °C (Jones 2005). Climate warming is shifting these bands toward higher latitude; traditional cool-climate regions such as Bordeaux, Burgundy, and the Mosel face earlier bud-break, compressed phenology, and alcohol inflation (Module 8), while new regions in southern England, Denmark, southern Sweden, Tasmania, southern Patagonia, and British Columbia have emerged as serious wine producers since ~2000 CE.

Interspecific Hybrids

In regions where downy and powdery mildew pressure, phylloxera, or severe winter cold make V. vinifera uneconomic, interspecific hybrids are grown: French–American hybrids (Baco Noir, Seyval Blanc, Vidal Blanc, Chambourcin) dating from 19th-century crosses with American species; eastern-European hybrids with V. amurensis (Zilga, Rondo); and newer “PIWI” cultivars (Regent, Solaris, Johanniter) bred in Germany and Switzerland to combine high disease resistance with vinifera-like flavour.

5. Aroma, Flavor, and Terpene Genetics

The grapevine genome encodes an unusually large family of terpene synthases (~70 members), reflecting the plant’s biochemical investment in volatiles used for pollinator attraction, herbivore defence, and, incidentally, human varietal character. Muscat cultivars carry a point mutation in the plastidial 1-deoxy-D-xylulose-5-phosphate synthase (VvDXS1, chr 5) that up-regulates the MEP pathway, flooding the berry with monoterpenes—principally linalool (floral, citrus), geraniol, nerol, and alpha-terpineol (Emanuelli 2010).

\[\text{MEP pathway: pyruvate + GAP} \xrightarrow{DXS} \text{DXP} \rightarrow \dots \rightarrow \text{GPP} \xrightarrow{\text{linalool synthase}} \text{linalool}\]

Methoxypyrazines and Cabernet Sauvignon

The “green pepper” character of under-ripe Cabernet Sauvignon and Sauvignon Blanc is caused by 3-isobutyl-2-methoxypyrazine (IBMP), a compound synthesised from leucine via the enzyme VvOMT3 (Dunlevy 2010). IBMP degrades photochemically in sun-exposed berries; hence the enological emphasis on canopy management for red Bordeaux varieties.

Thiols and Sauvignon Blanc

The tropical-fruit character of Marlborough (New Zealand) Sauvignon Blanc is driven by 3-mercaptohexan-1-ol (3MH) and 4-methyl-4-mercaptopentan-2-one (4MMP). These volatile thiols are released during fermentation from non-volatile cysteinylated and glutathionylated precursors in the grape must by yeast carbon–sulfur lyases (Dubourdieu 2006).

Rotundone and Syrah

The spicy, black-pepper character of cool-climate Syrah and Grüner Veltliner is due to a single sesquiterpene, rotundone, detectable at ~16 ng/L. Biosynthesis involves a cytochrome-P450-mediated oxidation of alpha-guaiene; genetic analysis localised the pathway to chr 8 but the specific structural gene remains the subject of active research.

6. Grapevine in the Post-Genomic Era

Since the 2007 Pinot Noir reference, the grape research community has released pangenomes spanning dozens of cultivars, long-read assemblies of Cabernet Sauvignon, Merlot, Sangiovese, Riesling, Chardonnay, and Muscat, plus high-density SNP arrays (Vitis 18K, 10K, GrapeReSeq 63K). Structural-variation surveys reveal that no two cultivars share more than ~97% of their sequence identity; highly heterozygous genomes routinely differ by hundreds of thousands of large structural variants.

CRISPR Editing and Disease Resistance

Because grapevine is vegetatively propagated, classical back-cross introgression from downy/powdery-mildew-resistant American and Asian species is slow and compromises the specific cultivar identity (“Cabernet Sauvignon cannot be anything else”). CRISPR editing of susceptibility alleles—for example VvMLO6/7/11 for powdery mildew, or VvDMR6 for downy mildew—offers a minimally invasive route to disease resistance that preserves the commercial phenotype (Wan 2019, Pessina 2016).

\[\text{Disease resistance} = \underbrace{\text{MLO knock-out}}_{\text{recessive}} \;\oplus\; \underbrace{\text{NLR stacking}}_{\text{dominant}}\]

Clonal Diversity Within Cultivars

Each of the famous vinifera cultivars is actually a small population of somatic clones accumulated over centuries of vegetative propagation. Hundreds of distinct Pinot Noir clones are registered in the French national clone catalogue (ENTAV-INRA); clone identity correlates with subtle differences in phenology, yield, and volatile profile, and has historically dwarfed inter-cultivar breeding in commercial importance.

6b. Cultivar Genealogy: Founders and Descendants

Parentage reconstruction using SSR and SNP markers has demonstrated that the most commercially important V. vinifera cultivars form a shallow pedigree of closely related individuals. Bowers & Meredith (1997) and Lacombe et al. (2013) resolved many famous parentages:

Cabernet Sauvignon = Cabernet Franc × Sauvignon Blanc.
Chardonnay = Pinot × Gouais Blanc.
Gamay = Pinot × Gouais Blanc.
Aligoté = Pinot × Gouais Blanc.
Auxerrois = Pinot × Gouais Blanc.
Melon de Bourgogne = Pinot × Gouais Blanc.
Merlot = Cabernet Franc × Magdeleine Noire des Charentes.
Syrah = Dureza × Mondeuse Blanche.
Pinot Gris = somatic bud-sport of Pinot Noir.
Pinot Blanc = bud-sport of Pinot Gris.
Zinfandel = Primitivo = Tribidrag (Croatian Crljenak Kaštelanski).
Cortese, Prosecco/Glera = relatives of Pinot lineage.

Gouais Blanc and Pinot are parental to at least 80 named cultivars across Europe; their central role in grape breeding is disproportionate to either variety’s present-day commercial importance (Gouais was banned in several French regions for low-quality wine!). Such shallow parentage reflects both the small founding stock of post-phylloxera European viticulture and the persistent tradition of selecting outstanding seedlings from chance crosses between a handful of well-planted cultivars.

7. Synthesis: Why Grape Biophysics?

Grape is the most economically valuable fruit crop on Earth (~$250 billion/yr of wine alone), and it is exceptional biophysically: a long-lived, woody, clonally propagated perennial with sophisticated secondary metabolism, precise climatic requirements, and millennia of human selection pressure. Course themes that flow from the evolutionary foundation of this module:

Berry development (Module 1): cell expansion, veraison, sugar loading.
Phenolics (Module 2): anthocyanins, tannins, proanthocyanidins; MybA1 and MybPA1/2 regulators.
Photosynthesis and ripening (Module 3): canopy light interception, assimilate partitioning.
Terroir and water stress (Module 4): vapour-pressure deficit, regulated deficit irrigation, soil effects on sugar/acid balance.
Sugar–acid balance (Module 5): invertase activity, malate decarboxylation, pH dynamics at harvest.
Pests and diseases (Module 6): phylloxera root biology, downy/powdery mildew, Pierce’s disease, grapevine trunk diseases.
Fermentation and wine biochemistry (Module 7): yeast redox, ester production, malolactic fermentation, oxidation pathways.
Climate change and viticulture (Module 8): phenological advance, alcohol inflation, AOC relocation, future cultivar design.

Module 0 gives the indispensable evolutionary frame: the bottleneck that limits genetic diversity, the phylloxera rescue that defines modern viticulture, the MybA1 locus that partitions world production into “white” and “red”, and the clonal propagation that freezes named cultivars into permanence.

Simulation 1: Coalescent Bottleneck of Grapevine Domestication

Fit a single-epoch bottleneck to the observed loss of heterozygosity between Caucasian wild V. vinifera ssp. sylvestris ($H \approx 0.78$) and modern cultivars ($H \approx 0.72$), and cross-check with a Monte Carlo coalescent simulation of TMRCA for a sample of 20 modern cultivars.

Python

script.py125 lines

import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

# Coalescent estimate of the Vitis vinifera domestication bottleneck.
# Grapevine was domesticated ~8000 BP in the Caucasus/Transcaucasus (McGovern 2003)
# from the wild dioecious progenitor Vitis vinifera ssp. sylvestris.
# Modern cultivated grapes (ssp. sativa) are overwhelmingly hermaphroditic.
#
# Observed heterozygosity in modern cultivars averages H_obs ~ 0.72 (Myles 2011)
# and in wild Caucasian sylvestris populations H_wild ~ 0.78.
# Under a single-epoch bottleneck of severity (N_b, duration T_b generations)
# followed by constant population size N_now, heterozygosity decays as:
#
#   H(t) = H_0 * (1 - 1/(2 N_b))^(T_b)   (bottleneck loss)
#   H_now = H_b                            (approximate, large N_now recovery)

H_wild = 0.78   # ancestral wild heterozygosity
H_obs  = 0.72   # modern cultivar heterozygosity

# generations since domestication (grapevine generation ~ 3 years; 8000 / 3 ~ 2700 gen)
T_b_grid = np.array([200, 500, 1000, 2000])   # bottleneck duration (gen)
Nb_grid  = np.linspace(50, 5000, 200)         # bottleneck effective size

fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(15, 11))
fig.patch.set_facecolor('#0a0a1a')

for ax in [ax1, ax2, ax3, ax4]:
    ax.set_facecolor('#0a0a1a')
    ax.tick_params(colors='#e9d5ff')
    for spine in ax.spines.values():
        spine.set_color('#581c87')
    ax.grid(True, color='#3b0764', alpha=0.35)

# ---- Panel 1: H vs N_b for several bottleneck durations ----
colors = ['#c084fc', '#d946ef', '#a855f7', '#7c3aed']
for Tb, col in zip(T_b_grid, colors):
    H = H_wild * (1 - 1/(2*Nb_grid))**Tb
    ax1.plot(Nb_grid, H, '-', color=col, linewidth=2.2, label=f'T_b = {Tb} gen')
ax1.axhline(H_obs, color='#fcd34d', linestyle='--', label=f'H_obs = {H_obs}')
ax1.set_xscale('log')
ax1.set_xlabel('Bottleneck effective size N_b', color='#e9d5ff', fontsize=11)
ax1.set_ylabel('Expected heterozygosity H', color='#e9d5ff', fontsize=11)
ax1.set_title('Heterozygosity decay vs. bottleneck size',
              color='#f3e8ff', fontsize=12, fontweight='bold')
ax1.legend(facecolor='#1e1b4b', edgecolor='#581c87', labelcolor='#e9d5ff', fontsize=9)

# ---- Panel 2: Best-fit N_b for fixed T_b = 2700 gen (grapevine from 8000 BP) ----
Tb = 2700
H_pred = H_wild * (1 - 1/(2*Nb_grid))**Tb
residual = np.abs(H_pred - H_obs)
Nb_fit = Nb_grid[np.argmin(residual)]
ax2.plot(Nb_grid, H_pred, '-', color='#a855f7', linewidth=2.2, label=f'Model T_b=2700 gen')
ax2.axhline(H_obs, color='#fcd34d', linestyle='--', linewidth=1.5, label='H_obs = 0.72')
ax2.axvline(Nb_fit, color='#f43f5e', linestyle=':', linewidth=1.5, label=f'N_b fit = {Nb_fit:.0f}')
ax2.set_xscale('log')
ax2.set_xlabel('N_b (log scale)', color='#e9d5ff', fontsize=11)
ax2.set_ylabel('H expected', color='#e9d5ff', fontsize=11)
ax2.set_title('Fit of domestication bottleneck size',
              color='#f3e8ff', fontsize=12, fontweight='bold')
ax2.legend(facecolor='#1e1b4b', edgecolor='#581c87', labelcolor='#e9d5ff', fontsize=9)

# ---- Panel 3: Monte Carlo coalescent simulation ----
rng = np.random.default_rng(2026)
n_samp = 20
N_wild = 50000
N_bott = Nb_fit
n_reps = 300

def coalesce_two_epoch(n, N1, N2, t_switch):
    # Backward simulation: lineages coalesce at rate k(k-1)/(2N). Switch N at t_switch.
    lineages = n
    t = 0.0
    tree_heights = []
    while lineages > 1:
        rate = lineages*(lineages-1) / 2.0
        if t < t_switch:
            dt = rng.exponential(1.0 / (rate / N2))
            if t + dt > t_switch:
                t = t_switch
                continue
        else:
            dt = rng.exponential(1.0 / (rate / N1))
        t += dt
        lineages -= 1
        tree_heights.append(t)
    return tree_heights[-1]

TMRCA = np.array([coalesce_two_epoch(n_samp, N_wild, N_bott, 2700) for _ in range(n_reps)])
ax3.hist(TMRCA / 1000, bins=30, color='#a855f7', edgecolor='#f3e8ff', alpha=0.85)
ax3.axvline(TMRCA.mean()/1000, color='#fcd34d', linestyle='--',
            label=f'Mean TMRCA = {TMRCA.mean():.0f} gen')
ax3.set_xlabel('TMRCA (thousands of generations)', color='#e9d5ff', fontsize=11)
ax3.set_ylabel('Count', color='#e9d5ff', fontsize=11)
ax3.set_title('Coalescent TMRCA distribution',
              color='#f3e8ff', fontsize=12, fontweight='bold')
ax3.legend(facecolor='#1e1b4b', edgecolor='#581c87', labelcolor='#e9d5ff', fontsize=9)

# ---- Panel 4: diversity trajectory through time ----
generations = np.arange(0, 3000)
H_traj = H_wild * np.where(generations < 2700,
                           (1 - 1/(2*Nb_fit))**generations,
                           (1 - 1/(2*Nb_fit))**2700)
ax4.plot(-generations, H_traj, '-', color='#a855f7', linewidth=2.5, label='H(t) under bottleneck')
ax4.fill_between(-generations, H_traj, alpha=0.3, color='#a855f7')
ax4.axhline(H_wild, color='#22d3ee', linestyle=':', label=f'H_wild = {H_wild}')
ax4.axhline(H_obs, color='#fcd34d', linestyle='--', label=f'H_obs = {H_obs}')
ax4.set_xlabel('Generations from present', color='#e9d5ff', fontsize=11)
ax4.set_ylabel('Heterozygosity', color='#e9d5ff', fontsize=11)
ax4.set_title('Heterozygosity trajectory during domestication',
              color='#f3e8ff', fontsize=12, fontweight='bold')
ax4.legend(facecolor='#1e1b4b', edgecolor='#581c87', labelcolor='#e9d5ff', fontsize=9)

# Integrated diversity loss using trapezoid
H_loss = np.trapezoid(H_wild - H_traj, generations)
print(f"Integrated heterozygosity deficit (trapezoid): {H_loss:.2f} het-gen")

plt.tight_layout()
plt.savefig('output.png', dpi=120, bbox_inches='tight', facecolor='#0a0a1a')

print(f"Best-fit bottleneck effective size: N_b = {Nb_fit:.0f}")
print(f"Corresponds to ~{100 * (H_wild - H_obs)/H_wild:.1f}% loss of heterozygosity")
print(f"Myles 2011: 950 Vitis vinifera cultivars show only ~19% unique parentage chains")
print(f"Most cultivars are sibs / parent-offspring of <20 founder vines.")

Click Run to execute the Python code

Code will be executed with Python 3 on the server

Simulation 2: Mendelian Segregation of the MybA1 Berry-Color Locus

Simulate an F2 cross (Ww × Ww) at the MybA1 locus, check the 3 : 1 phenotypic ratio with a chi-square test, and summarise the global distribution of ~10,000 named V. vinifera cultivars by berry color.

Python

script.py156 lines

import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

# Mendelian segregation at the MybA1 berry-color locus of Vitis vinifera.
# Walker et al. (2007) showed that the red/black pigmentation of vinifera berries
# is controlled by a cluster of MYB transcription factors on chromosome 2.
# The key functional gene is VvMybA1. A Gret1 retrotransposon insertion in the
# promoter silences VvMybA1 -> white berries. Homozygous insertion = white.
# Heterozygous or homozygous wild-type = colored (anthocyanin-producing).
#
# Allele W = functional (dominant, colored); w = retro-silenced (recessive, white)
# Classical crosses:
#   WW x ww -> all Ww colored
#   Ww x Ww -> 1 WW : 2 Ww : 1 ww  -> 3 colored : 1 white
#   Ww x ww -> 1 Ww : 1 ww         -> 1 colored : 1 white (test cross)

# Simulate F2 segregation with finite sample size
rng = np.random.default_rng(42)
N_F2 = 200
# Ww x Ww produces zygote genotype probability (1/4 WW, 1/2 Ww, 1/4 ww)
probs = np.array([0.25, 0.5, 0.25])
labels = ['WW', 'Ww', 'ww']
draws = rng.choice(3, size=N_F2, p=probs)
counts = np.bincount(draws, minlength=3)

# Phenotype: colored if at least one W
pheno = np.array(['Colored' if g < 2 else 'White' for g in draws])
n_colored = np.sum(pheno == 'Colored')
n_white   = np.sum(pheno == 'White')

# Chi-square test against 3:1 expectation
exp_col = 0.75 * N_F2
exp_wh  = 0.25 * N_F2
chi2 = (n_colored - exp_col)**2 / exp_col + (n_white - exp_wh)**2 / exp_wh

# --- Global cultivar catalogue ---
# Number of cultivars classified by berry color (OIV 2020, Jaillon 2007)
cult = {
    'Noir (black/red)': 6500,
    'Blanc (white)':    3200,
    'Gris (grey/pink)':  500,
    'Rouge/Rose':        300,
}

# Famous cultivars
famous = {
    'Pinot Noir':        'WW/Ww',
    'Chardonnay':        'ww (Gret1/Gret1)',
    'Cabernet Sauvignon':'Ww',
    'Muscat Blanc':      'ww (linalool+)',
    'Grenache Noir':     'Ww',
    'Riesling':          'ww',
    'Pinot Gris':        'Ww (partial loss in berry skin)',
}

fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(15, 11))
fig.patch.set_facecolor('#0a0a1a')

# ---- Panel 1: F2 genotype frequencies (observed vs expected) ----
x = np.arange(3)
ax1.bar(x - 0.2, counts,               width=0.4, color='#a855f7', label='Observed',
        edgecolor='#f3e8ff', alpha=0.9)
ax1.bar(x + 0.2, probs * N_F2, width=0.4, color='#c084fc', label='Expected (1:2:1)',
        edgecolor='#f3e8ff', alpha=0.7)
ax1.set_xticks(x)
ax1.set_xticklabels(labels, color='#e9d5ff')
ax1.set_ylabel('Count (N=200)', color='#e9d5ff', fontsize=11)
ax1.set_title('F2 genotypic ratios at MybA1',
              color='#f3e8ff', fontsize=12, fontweight='bold')
ax1.legend(facecolor='#1e1b4b', edgecolor='#581c87', labelcolor='#e9d5ff', fontsize=9)

# ---- Panel 2: Phenotypic 3:1 ratio + chi2 result ----
ax2.bar([0], [n_colored], width=0.6, color='#7c2d12', edgecolor='#f3e8ff', label='Colored (obs)')
ax2.bar([1], [n_white],   width=0.6, color='#fde68a', edgecolor='#f3e8ff', label='White (obs)')
ax2.axhline(exp_col, color='#a855f7', linestyle='--', label='Expected colored (3/4)')
ax2.axhline(exp_wh,  color='#22d3ee', linestyle='--', label='Expected white (1/4)')
ax2.set_xticks([0, 1])
ax2.set_xticklabels(['Colored', 'White'], color='#e9d5ff')
ax2.set_ylabel('Count', color='#e9d5ff', fontsize=11)
ax2.set_title(f'Phenotypic ratio (chi^2 = {chi2:.2f}, df=1)',
              color='#f3e8ff', fontsize=12, fontweight='bold')
ax2.legend(facecolor='#1e1b4b', edgecolor='#581c87', labelcolor='#e9d5ff', fontsize=9)

# ---- Panel 3: world cultivar catalogue pie chart ----
wedges, texts, autotexts = ax3.pie(
    list(cult.values()),
    labels=list(cult.keys()),
    colors=['#4c1d95', '#fef3c7', '#fb7185', '#f97316'],
    autopct='%1.1f%%',
    startangle=90,
    wedgeprops=dict(edgecolor='#0a0a1a', linewidth=1.5),
    textprops=dict(color='#e9d5ff', fontsize=10),
)
for t in autotexts:
    t.set_color('#0a0a1a')
    t.set_fontweight('bold')
ax3.set_title('10,000 Vitis vinifera cultivars by berry colour',
              color='#f3e8ff', fontsize=12, fontweight='bold')

# ---- Panel 4: cumulative distribution / Punnett grid ----
# 4x4 Punnett square image for Ww x Ww
grid = np.array([[0, 1, 1, 2],
                 [1, 0, 2, 1],
                 [1, 2, 0, 1],
                 [2, 1, 1, 0]])
# Relabel so values 0,1,2 correspond to WW, Ww, ww count

P = np.array([['WW', 'Ww'],
              ['Ww', 'ww']])
rows = ['W (maternal)', 'w (maternal)']
cols = ['W (paternal)', 'w (paternal)']
colours = {'WW': '#4c1d95', 'Ww': '#a855f7', 'ww': '#fde68a'}

for i in range(2):
    for j in range(2):
        geno = P[i, j]
        ax4.add_patch(plt.Rectangle((j, 1-i), 1, 1, facecolor=colours[geno],
                                      edgecolor='#f3e8ff', linewidth=1.5))
        col_txt = '#0a0a1a' if geno == 'ww' else '#f3e8ff'
        ax4.text(j + 0.5, 1-i + 0.55, geno, ha='center', va='center',
                 fontsize=22, fontweight='bold', color=col_txt)
        pheno = 'Colored' if geno != 'ww' else 'White'
        ax4.text(j + 0.5, 1-i + 0.25, pheno, ha='center', va='center',
                 fontsize=10, color=col_txt)
ax4.set_xlim(-0.4, 2.4)
ax4.set_ylim(-0.2, 2.4)
ax4.set_xticks([0.5, 1.5]); ax4.set_xticklabels(cols, color='#e9d5ff')
ax4.set_yticks([1.5, 0.5]); ax4.set_yticklabels(rows, color='#e9d5ff')
ax4.set_title('Punnett square: Ww x Ww at MybA1',
              color='#f3e8ff', fontsize=12, fontweight='bold')
ax4.grid(False)

# integrate Hardy-Weinberg equilibrium Q(p)=2pq between allele frequency 0 and 1
ps = np.linspace(0, 1, 200)
heterozygosity = 2 * ps * (1 - ps)
HW_area = np.trapezoid(heterozygosity, ps)  # = 1/3
print(f"Hardy-Weinberg integral of 2pq over p=[0,1] (trapezoid): {HW_area:.3f}")

plt.tight_layout()
plt.savefig('output.png', dpi=120, bbox_inches='tight', facecolor='#0a0a1a')

print(f"Simulated F2: WW={counts[0]}, Ww={counts[1]}, ww={counts[2]}")
print(f"Phenotypic ratio {n_colored}:{n_white} (expected 3:1 = {int(exp_col)}:{int(exp_wh)})")
print(f"Chi^2 = {chi2:.2f} (critical 3.84 at alpha=0.05, df=1)")
print("Famous cultivars at MybA1:")
for name, geno in famous.items():
    print(f"  {name}: {geno}")

Click Run to execute the Python code

Code will be executed with Python 3 on the server

Key References

• Jaillon, O. et al. (2007). “The grapevine genome sequence suggests ancestral hexaploidization in major angiosperm phyla.” Nature, 449, 463–467.

• McGovern, P. E. (2003). Ancient Wine: The Search for the Origins of Viniculture. Princeton University Press.

• Myles, S. et al. (2011). “Genetic structure and domestication history of the grape.” PNAS, 108, 3530–3535.

• Walker, A. R. et al. (2007). “White grapes arose through the mutation of two similar and adjacent regulatory genes.” Plant Journal, 49, 772–785.

• Ramos-Madrigal, J. et al. (2019). “Palaeogenomic insights into the origins of French grapevine diversity.” Nature Plants, 5, 595–603.

• Dong, Y. et al. (2023). “Dual domestications and origin of traits in grapevine evolution.” Science, 379, 892–901.

• Picq, S. et al. (2014). “A small XY chromosomal region explains sex determination in wild dioecious V. vinifera and the reversal to hermaphroditism in domesticated grapevines.” BMC Plant Biology, 14, 229.

• Massonnet, M. et al. (2020). “The genetic basis of sex determination in grapes.” Nature Communications, 11, 2902.

• Granett, J., Walker, M. A., Kocsis, L. & Omer, A. D. (2001). “Biology and management of grape phylloxera.” Annual Review of Entomology, 46, 387–412.

• Kobayashi, S., Goto-Yamamoto, N. & Hirochika, H. (2004). “Retrotransposon-induced mutations in grape skin color.” Science, 304, 982.

← Course Overview Module 1: Berry Development →