Genomic epidemiology of hospital related infections in Hong Kong 2022

This project focuses on analyzing hospital-related infections in Hong Kong, including case classifications, genomic surveillance, and mobility data. The repository contains scripts, data, and results for visualizing and interpreting the findings.

Project Structure

1. Data

The data/ directory contains raw and processed data files used in the analysis. Key subdirectories include:

• case_curve/: Contains JSON and Excel files for aggregated case data.
  • HK_case_data.json: JSON file with aggregated confirmed case classifications.
  • hk_epicurve.xlsx: Processed case data in Excel format.
• hospital_data/: Contains metadata for hospitals and districts.
  • metadata_cases_plot.csv: Metadata for hospital-acquired infections.
• mobility_data/: Contains mobility and transport data.
  • 2022_HK_Region_Mobility_Report.csv: Google Community Mobility Report for Hong Kong.
  • statistics_on_daily_passenger_traffic.csv: Cross-border passenger traffic data.
  • public_transport.xlsx: Local transport data.

2. Scripts

The scripts/ directory contains R scripts for data processing and visualization. Key scripts include:

1. metadata_visualization.R

   • Reads and processes metadata from various sources.
   • Generates figures and summarizing tables about case data, such as Panel A/B epidemiological curves.

2. tree_visualization.R

   • Loads and processes MCC (BEAST) and ML phylogenetic trees.
   • Annotates hospital clusters, collapses large community clades, and creates publication-quality tree figures.

3. GAM.R

   • Reads Markov jump data and predictor datasets.
   • Builds and diagnoses a GAM (Negative Binomial) model, performs VIF checks, and creates time-series plots of residual autocorrelation.

3. Results

The results/ directory contains output figures and processed data

How to Run

1. Install Required Libraries: Ensure the following R libraries are installed:

   • tidyverse, jsonlite, lubridate, ggplot2, ggforce, ggrepel, MetBrewer, ggbump, scales, sp, gridExtra, RColorBrewer, readxl, writexl.

2. Run Scripts:

   • Execute 1.metadata_visualization.R to generate Figures 1a, 1b, 1c, and 1d.
   • Execute 2.tree_visualization.R to generate phylogenetic tree visualizations.
   • Execute 3.GAM.R to generate GAM results.
   • Execute GISAID_data_processing/QC.R to process and filter GISAID genomic data. (Details refer to here)

3. View Results:

   • Figures are saved in the results/ directory.

Citation

Pending

---