COVID-19  Patient Network

- The project analyzed network characteristics of COVID-19 patients to identify factors associated with mortality and recovery.

- Network construction linked patients (nodes) based on similarity thresholds.

- Community detection was performed using label propagation, with modularity values around 0.4 for all networks.

- Key findings for the deceased patient network:
- Important features: Smoking, Race, Hypertension, Diabetes
- Strongly connected with heavy-tailed degree distribution

- Key findings for the recovered patient network:
- Important features: Race, Heart failure, Age, Gender, Hypertension

- The combined network showed similar important features to the deceased network, with the addition of Age.

- All networks exhibited strong connectivity and heavy-tailed degree distributions.

- Robustness analysis revealed:
- Critical threshold (Fc) of 0.5, indicating vulnerability to targeted attacks
- Targeting high-degree nodes could significantly disrupt the network

- Potential intervention strategy: Target patients with specific characteristics (Hypertension, Diabetes, Male, White, Non-Smoking) to reduce mortality.

- The project demonstrated the effectiveness of network analysis in understanding COVID-19 patient characteristics and identifying potential intervention strategies.