Proteomic Organ Age Clock

This example loads proteomic data and runs it through the OrganAge clock

First load the COVID-19 serum proteomics data using the data library

from biolearn.data_library import DataLibrary
data_source = DataLibrary().get("Covid19_Filbin_Plasma")
data = data_source.load()

Now run the organ age clocks on the proteomic data

from biolearn.model_gallery import ModelGallery
gallery = ModelGallery()
chronological_results = gallery.get("OrganAge1500Chronological").predict(data)

Prepare data for visualization

import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

plot_data_dict = {'age_range': data.metadata['age_range']}

ORGANS = ['Organismal', 'Multi-organ', 'Brain', 'Heart', 'Immune']
for organ in ORGANS:
    plot_data_dict[organ] = chronological_results[organ]

plot_data = pd.DataFrame(plot_data_dict)
plot_data = plot_data.dropna(subset=['age_range'])

Create boxplot visualization

sns.set_style("whitegrid")
fig, ax = plt.subplots(figsize=(12, 7))

# Sort age ranges properly
age_order = sorted(plot_data['age_range'].unique())

plot_data_long = plot_data.melt(
    id_vars=['age_range'],
    value_vars=ORGANS,
    var_name='Organ',
    value_name='Predicted_Age'
)
sns.boxplot(
    data=plot_data_long,
    x='age_range',
    y='Predicted_Age',
    hue='Organ',
    order=age_order,
    ax=ax
)


ax.set_xlabel('Chronological Age Range', fontsize=12)
ax.set_ylabel('Predicted Organ Age', fontsize=12)
ax.set_title('Organ Age Predictions by Age Group', fontsize=14)
ax.legend(title='Organ', bbox_to_anchor=(1.05, 1), loc='upper left')
plt.tight_layout()
plt.show()

Estimated memory usage: 184 MB