plot_structure_similarity_metrics_AF3.py

from util import *
import seaborn as sns
import matplotlib.pyplot as plt
import pandas as pd
from matplotlib.gridspec import GridSpec
from mpl_toolkits.axes_grid1 import make_axes_locatable

import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.gridspec import GridSpec
import os

csv_file = './CSPRANK.csv'
x_col = 'AF3_DockQ_top_rank'
y_col = 'AF3_TM'

# Read data from CSV
df = pd.read_csv(csv_file)

fig = plt.figure(figsize=(8, 8))
gs = GridSpec(4, 4, figure=fig)

# Axes definitions
ax_scatter = fig.add_subplot(gs[1:4, 0:3])
ax_histx   = fig.add_subplot(gs[0, 0:3], sharex=ax_scatter)
ax_histy   = fig.add_subplot(gs[1:4, 3], sharey=ax_scatter)

# Scatter plot
ax_scatter.scatter(df[x_col], df[y_col], alpha=0.6, edgecolors='k')
ax_scatter.set_xlabel('AF3 DockQ')
ax_scatter.set_ylabel('AF3 TM')
ax_scatter.set_xlim(0, 1)
ax_scatter.set_ylim(0, 1)

# Add dashed red lines
ax_scatter.axhline(y=0.8, color='red', linestyle='--')
ax_scatter.axvline(x=0.4, color='red', linestyle='--')

# Shade the region x > 0.4 and y > 0.8 a light gray
ax_scatter.fill_betweenx(y=[0.8, 1], x1=0.4, x2=1, color='gray', alpha=0.3)

# Histograms
ax_histx.hist(df[x_col], bins=30, alpha=0.7, color='blue')
ax_histy.hist(df[y_col], bins=30, orientation='horizontal', alpha=0.7, color='blue')

# Turn off tick labels on hist plots for clarity
plt.setp(ax_histx.get_xticklabels(), visible=False)
plt.setp(ax_histy.get_yticklabels(), visible=False)

# Adjust spacing around subplots to make space for labels
plt.subplots_adjust(hspace=0.05, wspace=0.05)

# Save the plot
output_dir = './Figures'
os.makedirs(output_dir, exist_ok=True)
output_file = os.path.join(output_dir, 'structure_similarity_metrics_plot.png')
plt.savefig(output_file)



# Calculate the percentage of entries with AF3_TM > 0.8
percent_AF3_TM_gt_0_8 = (df['AF3_TM'] > 0.8).mean() * 100
print(f"Percentage of entries with AF3_TM > 0.8: {percent_AF3_TM_gt_0_8:.2f}%")

# Calculate the percentage of entries with AF3_DockQ_top_rank > 0.4
percent_AF3_DockQ_gt_0_4 = (df['AF3_DockQ_top_rank'] > 0.4).mean() * 100
print(f"Percentage of entries with AF3_DockQ_top_rank > 0.4: {percent_AF3_DockQ_gt_0_4:.2f}%")

# Calculate the percentage of entries with both conditions
percent_both_conditions = ((df['AF3_TM'] > 0.8) & (df['AF3_DockQ_top_rank'] > 0.4)).mean() * 100
print(f"Percentage of entries with both AF3_TM > 0.8 and AF3_DockQ_top_rank > 0.4: {percent_both_conditions:.2f}%")