feat(pygal): implement manhattan-gwas (#2972)

## Implementation: `manhattan-gwas` - pygal

Implements the **pygal** version of `manhattan-gwas`.

**File:** `plots/manhattan-gwas/implementations/pygal.py`

**Parent Issue:** #2925

---
:robot: *[impl-generate
workflow](https://github.com/MarkusNeusinger/pyplots/actions/runs/20612794558)*

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

Author: github-actions[bot]

plots/manhattan-gwas/implementations/pygal.py

@@ -0,0 +1,218 @@
+""" pyplots.ai
+manhattan-gwas: Manhattan Plot for GWAS
+Library: pygal 3.1.0 | Python 3.13.11
+Quality: 82/100 | Created: 2025-12-31
+"""
+
+import numpy as np
+import pygal
+from pygal.style import Style
+
+
+# Seed for reproducibility
+np.random.seed(42)
+
+# Chromosome lengths (simplified, in Mb scale)
+chrom_lengths = {
+    "1": 249,
+    "2": 243,
+    "3": 198,
+    "4": 191,
+    "5": 182,
+    "6": 171,
+    "7": 159,
+    "8": 146,
+    "9": 141,
+    "10": 136,
+    "11": 135,
+    "12": 134,
+    "13": 115,
+    "14": 107,
+    "15": 102,
+    "16": 90,
+    "17": 83,
+    "18": 80,
+    "19": 59,
+    "20": 64,
+    "21": 47,
+    "22": 51,
+}
+
+chromosomes = list(chrom_lengths.keys())
+n_snps_per_chrom = 500
+
+# Storage for data
+all_chroms = []
+all_cumulative_pos = []
+all_pvalues = []
+
+# Track chromosome boundaries for x-axis labels
+chrom_midpoints = []
+cumulative_offset = 0
+
+for chrom in chromosomes:
+    length = chrom_lengths[chrom]
+    positions = np.sort(np.random.uniform(0, length, n_snps_per_chrom))
+
+    # Store chromosome midpoint for x-axis label
+    chrom_midpoints.append(cumulative_offset + length / 2)
+
+    # Generate p-values: mostly uniform with some significant hits
+    pvalues = np.random.uniform(0, 1, n_snps_per_chrom)
+
+    # Add significant peaks on selected chromosomes
+    if chrom in ["6", "11", "16"]:
+        n_sig = np.random.randint(3, 8)
+        sig_indices = np.random.choice(n_snps_per_chrom, n_sig, replace=False)
+        pvalues[sig_indices] = 10 ** (-np.random.uniform(8, 15, n_sig))
+
+    # Add suggestive signals
+    n_sugg = np.random.randint(5, 15)
+    sugg_indices = np.random.choice(n_snps_per_chrom, n_sugg, replace=False)
+    pvalues[sugg_indices] = 10 ** (-np.random.uniform(5, 8, n_sugg))
+
+    cumulative_positions = positions + cumulative_offset
+    all_chroms.extend([chrom] * n_snps_per_chrom)
+    all_cumulative_pos.extend(cumulative_positions)
+    all_pvalues.extend(pvalues)
+
+    cumulative_offset += length
+
+# Convert to -log10 p-values
+neg_log_pvalues = -np.log10(np.array(all_pvalues))
+
+# Thresholds
+genome_wide_sig = -np.log10(5e-8)  # ~7.3
+suggestive_sig = 5.0  # -log10(1e-5)
+
+# Custom style with explicit colors for threshold lines
+custom_style = Style(
+    background="white",
+    plot_background="white",
+    foreground="#333333",
+    foreground_strong="#333333",
+    foreground_subtle="#666666",
+    colors=(
+        "#306998",  # Blue - odd chromosomes
+        "#888888",  # Gray - even chromosomes
+        "#D62728",  # Red - significant points above threshold
+        "#D62728",  # Red - genome-wide threshold line
+        "#FF7F0E",  # Orange - suggestive threshold line
+    ),
+    title_font_size=56,
+    label_font_size=36,
+    major_label_font_size=32,
+    legend_font_size=32,
+    value_font_size=20,
+    font_family="sans-serif",
+)
+
+# Create XY scatter chart
+chart = pygal.XY(
+    width=4800,
+    height=2700,
+    style=custom_style,
+    title="manhattan-gwas · pygal · pyplots.ai",
+    x_title="Chromosome",
+    y_title="-log₁₀(p-value)",
+    show_legend=True,
+    legend_at_bottom=True,
+    legend_at_bottom_columns=5,
+    legend_box_size=24,
+    stroke=False,
+    dots_size=5,
+    show_x_guides=False,
+    show_y_guides=True,
+    truncate_label=-1,
+    print_values=False,
+    x_label_rotation=0,
+    range=(0, 16),
+    xrange=(0, cumulative_offset),
+    tooltip_border_radius=10,
+    explicit_size=True,
+    spacing=30,
+    margin=60,
+    margin_bottom=200,
+    margin_top=100,
+)
+
+# Create x_labels with chromosome numbers at midpoints
+# Use string labels for chromosomes 1-22
+chart.x_labels = [str(i + 1) for i in range(len(chromosomes))]
+chart.x_labels_major_count = len(chromosomes)
+
+# Map positions to display label indices
+# We need to normalize x values to show chromosome labels
+# Pygal XY chart needs numeric x_labels for scatter, so we'll use custom formatter
+label_positions = chrom_midpoints.copy()
+
+
+def format_x_label(x_val):
+    """Find closest chromosome midpoint and return chromosome number."""
+    if not label_positions:
+        return ""
+    min_dist = float("inf")
+    closest_idx = 0
+    for i, pos in enumerate(label_positions):
+        dist = abs(x_val - pos)
+        if dist < min_dist:
+            min_dist = dist
+            closest_idx = i
+    # Only return label if close to midpoint (within half chromosome width)
+    if min_dist < 50:
+        return str(closest_idx + 1)
+    return ""
+
+
+# Set numeric x_labels at chromosome midpoints
+chart.x_labels = chrom_midpoints
+chart.x_value_formatter = lambda x: format_x_label(x) if x in chrom_midpoints else ""
+
+# Prepare data by chromosome with alternating colors
+odd_chrom_points = []
+even_chrom_points = []
+significant_points = []  # Points above genome-wide significance
+
+for idx, chrom in enumerate(chromosomes):
+    chrom_mask = [c == chrom for c in all_chroms]
+    chrom_x = [all_cumulative_pos[j] for j in range(len(all_chroms)) if chrom_mask[j]]
+    chrom_y = [neg_log_pvalues[j] for j in range(len(all_chroms)) if chrom_mask[j]]
+
+    for x, y in zip(chrom_x, chrom_y, strict=True):
+        point = {"value": (x, y), "label": f"Chr {chrom}: {x:.1f} Mb, -log₁₀(p)={y:.2f}"}
+        if y >= genome_wide_sig:
+            significant_points.append(point)
+        elif idx % 2 == 0:
+            odd_chrom_points.append(point)
+        else:
+            even_chrom_points.append(point)
+
+# Add chromosome data series (blue for odd, gray for even)
+chart.add("Odd chromosomes", odd_chrom_points, stroke=False, show_dots=True)
+chart.add("Even chromosomes", even_chrom_points, stroke=False, show_dots=True)
+
+# Add significant points as separate series (highlighted)
+chart.add("Significant (p<5×10⁻⁸)", significant_points, stroke=False, show_dots=True)
+
+# Add threshold lines as dense scatter points (works better in PNG than stroke)
+# Using many small dots to create visible line effect
+n_line_points = 200
+threshold_x = np.linspace(10, cumulative_offset - 10, n_line_points)
+
+# Genome-wide significance threshold line (y ≈ 7.3)
+gw_line_points = [
+    {"value": (x, genome_wide_sig), "label": f"Genome-wide threshold: -log₁₀(5×10⁻⁸) = {genome_wide_sig:.1f}"}
+    for x in threshold_x
+]
+chart.add("p = 5×10⁻⁸ threshold", gw_line_points, stroke=True, show_dots=True, dots_size=2)
+
+# Suggestive threshold line (y = 5)
+sugg_line_points = [
+    {"value": (x, suggestive_sig), "label": f"Suggestive threshold: -log₁₀(1×10⁻⁵) = {suggestive_sig:.1f}"}
+    for x in threshold_x
+]
+chart.add("p = 1×10⁻⁵ threshold", sugg_line_points, stroke=True, show_dots=True, dots_size=2)
+
+# Save outputs
+chart.render_to_file("plot.html")
+chart.render_to_png("plot.png")

plots/manhattan-gwas/metadata/pygal.yaml

@@ -0,0 +1,29 @@
+library: pygal
+specification_id: manhattan-gwas
+created: '2025-12-31T05:37:41Z'
+updated: '2025-12-31T15:21:52Z'
+generated_by: claude-opus-4-5-20251101
+workflow_run: 20612794558
+issue: 2925
+python_version: 3.13.11
+library_version: 3.1.0
+preview_url: https://storage.googleapis.com/pyplots-images/plots/manhattan-gwas/pygal/plot.png
+preview_thumb: https://storage.googleapis.com/pyplots-images/plots/manhattan-gwas/pygal/plot_thumb.png
+preview_html: https://storage.googleapis.com/pyplots-images/plots/manhattan-gwas/pygal/plot.html
+quality_score: 82
+review:
+  strengths:
+  - Excellent alternating blue/gray color scheme for chromosome distinction that is
+    colorblind-friendly
+  - Both genome-wide significance (p=5×10⁻⁸) and suggestive (p=1×10⁻⁵) threshold lines
+    are clearly visible
+  - Good use of pygal's tooltip system with informative labels showing chromosome,
+    position, and p-value
+  - Significant SNPs are properly highlighted in red as a separate series
+  - Clean chromosome labeling at midpoints with all 22 chromosomes visible
+  weaknesses:
+  - Contains a helper function (format_x_label) which violates the KISS principle
+    requiring flat script structure
+  - Data point markers are somewhat small for the 11,000+ point dataset; larger dots_size
+    would improve visibility
+  - Legend at bottom takes substantial vertical space; could be more compact