Bio machine learning survival analysis

Skill bg-szy/TOP-SKILLS/skills/awesome-skills/bio-machine-learning-survival-analysis

全球最大的 Claude Code 技能聚合库 · 收录 3900+ 来自 12+ 来源的技能，提供在线搜索与趋势分析看板 / The world's largest Claude Code skill aggregation hub — 3900+ skills from 12+ sources with online search and trend dashboard

Install

npx -y skills add bg-szy/TOP-SKILLS --skill bio-machine-learning-survival-analysis

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
4 stars4 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Analyzes time-to-event data using Kaplan-Meier curves, log-rank tests, and Cox proportional hazards regression with lifelines. Builds survival models from clinical and omics features. Use when predicting patient survival or modeling time-to-event outcomes.

SKILL.md

4.7 KB, as published. Nobody here has run it

Survival Prediction with lifelines

Kaplan-Meier Curves

from lifelines import KaplanMeierFitter
import matplotlib.pyplot as plt

kmf = KaplanMeierFitter()

# T: time to event or censoring
# E: event indicator (1=event occurred, 0=censored)
kmf.fit(T, event_observed=E)

# Plot survival curve
kmf.plot_survival_function()
plt.xlabel('Time (months)')
plt.ylabel('Survival probability')
plt.savefig('km_curve.png', dpi=150)

Compare Groups with Log-Rank Test

from lifelines import KaplanMeierFitter
from lifelines.statistics import logrank_test
import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(8, 6))

for group, color in zip(['high', 'low'], ['red', 'blue']):
    mask = df['risk_group'] == group
    kmf = KaplanMeierFitter()
    kmf.fit(df.loc[mask, 'time'], event_observed=df.loc[mask, 'event'], label=group)
    kmf.plot_survival_function(ax=ax, color=color)

# Log-rank test
high = df[df['risk_group'] == 'high']
low = df[df['risk_group'] == 'low']
results = logrank_test(high['time'], low['time'], event_observed_A=high['event'], event_observed_B=low['event'])
print(f'Log-rank p-value: {results.p_value:.4e}')

ax.set_xlabel('Time (months)')
ax.set_ylabel('Survival probability')
ax.set_title(f'Log-rank p = {results.p_value:.4e}')
plt.savefig('km_comparison.png', dpi=150)

Cox Proportional Hazards Regression

from lifelines import CoxPHFitter

# Prepare data: must have 'time' and 'event' columns
# Include covariates as additional columns
cph = CoxPHFitter()
cph.fit(df, duration_col='time', event_col='event')

# Summary with hazard ratios
cph.print_summary()

# Get hazard ratios as DataFrame
hr = cph.summary[['exp(coef)', 'exp(coef) lower 95%', 'exp(coef) upper 95%', 'p']]
print(hr)

# Concordance index (c-index): 0.5=random, 1.0=perfect
print(f'C-index: {cph.concordance_index_:.3f}')

Multivariate Cox Model

from lifelines import CoxPHFitter
import pandas as pd

# Combine clinical and omics features
cox_df = pd.DataFrame({
    'time': meta['survival_months'],
    'event': meta['vital_status'],
    'age': meta['age'],
    'stage': meta['stage_numeric'],
    'GENE1': expr.loc['GENE1'],
    'GENE2': expr.loc['GENE2']
})

cph = CoxPHFitter(penalizer=0.1)  # L2 regularization for stability
cph.fit(cox_df, duration_col='time', event_col='event')
cph.print_summary()

Predict Risk Scores

# Partial hazard (risk score)
risk_scores = cph.predict_partial_hazard(cox_df)

# Median risk split for KM plot
df['risk_group'] = (risk_scores > risk_scores.median()).map({True: 'high', False: 'low'})

Check Proportional Hazards Assumption

# Test PH assumption
cph.check_assumptions(df, p_value_threshold=0.05, show_plots=True)

Survival at Specific Time

# Survival probability at specific times
survival_probs = kmf.survival_function_at_times([12, 24, 60])
print(survival_probs)

# Median survival
print(f'Median survival: {kmf.median_survival_time_:.1f}')

Feature Selection for Survival

from lifelines import CoxPHFitter
import pandas as pd

# Univariate screening
results = []
for gene in expr.index[:1000]:
    cox_df = pd.DataFrame({
        'time': meta['survival_months'],
        'event': meta['vital_status'],
        'gene': expr.loc[gene]
    })
    cph = CoxPHFitter()
    cph.fit(cox_df, duration_col='time', event_col='event')
    results.append({
        'gene': gene,
        'hr': cph.hazard_ratios_['gene'],
        'p': cph.summary.loc['gene', 'p']
    })

results_df = pd.DataFrame(results)
sig_genes = results_df[results_df['p'] < 0.05].sort_values('p')

Related Skills

clinical-databases/variant-prioritization - Clinical variant interpretation
differential-expression/de-results - Find DE genes for survival model
machine-learning/biomarker-discovery - Select predictive features