SARS-CoV-2 Recombination Analysis

Table of Contents

• Setup
• Preparing Data for SARS-CoV-2 Recombination Analysis
  • Fetching UShER Mutation-Annotated Tree
  • Post-Processing RIVET Outputs
  • Running RIVET to infer recombination
  • Generating Data Files
• Notebook for Recombination Analysis
  • Supplemental Analysis
• References

Setup

This repository uses the pixi package manager. Install pixi if not already installed on your system with the following command. If you already have pixi installed on your system, please skip to the step below to activate the pixi shell environment.

NOTE: This command will work for Linux & macOS installation only, please see the pixi installation page for Windows: install pixi.

curl -fsSL https://pixi.sh/install.sh | sh

To clone this repository, type the following command: (This will take a couple of minutes)

git clone --recurse-submodules https://github.com/TurakhiaLab/SARS2_RecombinationAnalysis.git
cd SARS2_RecombinationAnalysis

Preparing Data for SARS-CoV-2 Recombination Analysis

The recombination analysis requires the following files, available in the data directory. You can either use these files directly and proceed to the notebook, or generate your own by following the steps below.

• substitutions_scores.csv: A file containing the PyR0 fitness scores for all ranked substitutions considered by PyR0 found in the "2023-12-25" private MAT.

• monthly_fitness_stats.csv: A file containing some basic statistics for the circulating fitness of all SARS-CoV-2 samples found in the MAT for each month.

• rivet_recombs_data.csv: The file containing all the information and statistics for the detected recombinants included in this analysis.

• genetic-diversity-gisaidAndPublic.2023-12-25.csv: The computed standing genetic diversity scores for the "2023-12-25" private MAT.

• reference.fasta: The SARS-CoV-2 reference genome.

The following two files will be automatically downloaded or generated and placed into the data directory when the pixi run data command is run:

• time_series_covid19_confirmed_global.csv: The JHU SARS-CoV-2 global daily cumulative confirmed case count data.
• mutations.tsv: Contains the individual amino acid mutation fitness scores from the PyR0 model, ranked by statistical significance.
• chronumental_dates_gisaidAndPublic.2023-12-25-STEPS2000-SERIAL3.metadata.tsv.tsv: The Chronumental-inferred emergence dates for all samples in the MAT.

Fetching UShER Mutation-Annotated Tree

A private MAT (dated "2023-12-25") containing GISAID sequences with privileged access was used for the analysis in this manuscript. Please feel free to email us if you need access to this MAT and its metadata file. The full analysis is also completely compatible with the publicly available MATs that can be downloaded from: https://hgdownload.soe.ucsc.edu/goldenPath/wuhCor1/UShER_SARS-CoV-2/. To reproduce this analysis with any MAT, replace relevant fields (shown below) in the config.yaml file with the correct name and date of the MAT and metadata files used. Make sure these files are copied into the data directory.

# Date of the MAT used for analysis
MAT_DATE: "2023-12-25"
# Name of the MAT and metadata file used in this analysis
MAT: gisaidAndPublic.2023-12-25.masked.nextclade.pangolin.pb
METADATA: gisaidAndPublic.2023-12-25.metadata.tsv

Post-Processing RIVET Outputs

If you have already run RIVET separately, copy the following two RIVET output files into the data directory. Otherwise, first run RIVET separately, and then copy these two output files into the data directory before proceeding to the next steps. For instructions on running RIVET, please see the section below titled: Running RIVET to infer recombination.

• final_recombinants_2023-12-25.txt: The RIVET inferred recombination results from the "2023-12-25" private MAT.
• trios_2023-12-25.vcf: The VCF file for all trio recombinant sequences inferred by RIVET.

cp final_recombinants_2023-12-25.txt data/
cp trios_2023-12-25.vcf data/

Running RIVET to infer recombination

Please see the instructions for installing RIVET on your machine: Use RIVET Locally

Once RIVET is installed, please see the instructions for running the RIVET "backend" pipeline and the required inputs to infer recombination: RIVET Backend Input

Two of the RIVET output files, final_recombinants_<DATE>.txt and trios.vcf were used in this analysis (RIVET_RESULTS_FILE and RIVET_VCF_FILE fields in config.yaml).

Generating Data Files

Follow these steps to generate all the data files used in the recombination analysis notebook, primarily rivet_recombs_data.csv.

Chronumental will be automatically re-run, since the inferred emergence dates of recombinants are required for this analysis. See the config.yaml field below.

RERUN_CHRONUMENTAL: True

Run the following command to generate the data used in this analysis:

pixi run data

If you are using a different MAT than the one used in this analysis or wish to re-generate these results (already included in the data directory for the MAT used in this analysis), follow the instructions at the link provided below to reproduce the entire standing genetic diversity file (genetic-diversity-gisaidAndPublic.2023-12-25.csv) for all months.

• Instructions: Calculate Standing Genetic Diversity

If you want to reproduce the monthly circulating fitness statistics (monthly_fitness_stats.csv), you can run the pixi task command below, after running the above pixi run data command to generate the Chronumental file. This command requires the Chronumental dates output file to be in the data directory.

pixi run circulating-fitness-stats

Notebooks for Recombination Analysis

The Jupyter notebook notebooks/analysis.ipynb reproduces the analyses and statistics reported in the manuscript using the following files from the data directory:

• rivet_recombs_data.csv
• substitutions_scores.csv
• monthly_fitness_stats.csv

Run the following two commands to launch the notebooks/analysis.ipynb notebook.

# First, activate the shell environment that contains all the necessary dependencies
pixi shell
# Then launch the notebook
pixi run jupyter lab notebooks/analysis.ipynb

Supplemental Analysis

For more documentation on reproducing the supplemental analysis and figures, please see this section: Supplemental Analysis

References

This work was only made possible by the important contributions of the following methods and datasets. In addition to this manuscript, please consider citing the following papers if you found this analysis useful for your research.

• PyR0 Model

  • Obermeyer, F. et al. Analysis of 6.4 million SARS-CoV-2 genomes identifies mutations associated with fitness. Science 376, 1327–1332 (2022).

• JHU CSSE COVID-19 Dataset

  • Dong, E., Du, H. & Gardner, L. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect. Dis. 20, 533–534 (2020).

• Chronumental

  • Sanderson, T. Chronumental: time tree estimation from very large phylogenies. 2021.10.27.465994 Preprint at https://doi.org/10.1101/2021.10.27.465994 (2021).