Created for the JEV Austrakka Portal.

This repository contains 3 snakemake workflows for the analysis of JEV virus data:

1. ingest/ - Download data from GenBank, clean and curate it
2. nextclade_data/ - Create tree.json for nextclade dataset
3. phylogenetic/ - Make phylogenetic trees for each genotype

mamba env create -f conda.yaml
conda activate jev-inp

cd ingest
snakemake -j 8

There are three options:

1. Full workflow: Each pipeline must be run separately and generally in this order ingest -> nextclade_data -> phylogenetic. The output of the final pipeline is an annotated, ancestrally reconstructed (.json) phylotree for each JEV genotype. Use auspice.us to view the trees.

2. Update Nextclade dataset: A nextclade dataset was created by running ingest -> nextclade_data to generate a nextclade dataset for genotyping and qcing. Unless large amounts of genomic data (eg for JEV5) are uploaded, there is no need to rerun these steps.

3. Create Phylogenies with new (user) sequences:

• You only need to run the phylogenetics workflow. Put all your sequences into a single .fasta file including different JEV genotypes. The pipeline QC and genotype your data then split as required.

• Modify the austrakka value in phylogenetics/config/config.yaml to point towards your new sequences. You will need 2 files: all.fasta and all.tsv. all.tsv needs to contain the following columns: Seq_ID, date, region, Country, Host. See phylogenetics/resources/example/ for an example.

datasets:
 nextclade: "../nextclade_dataset/resources/nc_dataset/"
 backbone: "resources/backbone/"
 austrakka: "resources/examples/" # <- change this path

• Activate conda env:

conda activate jev-inp

• enter the phylogenetic directory and run the snakemake workflow:

snakemake -j 8

• The output of the pipeline will be in output/auspice/:

jev1.json
jev1_root-sequence.json
jev2.json
jev2_root-sequence.json
jev3.json
jev3_root-sequence.json
jev4.json
jev4_root-sequence.json
jev5.json
jev5_root-sequence.json

• Drag and drop any of the .json files into auspice.us (ignore the root-sequence.json.) to visualise.
• If you need the newick format, it will be in output/jev{genotype}/tree_jev{genotype}.nwk

• Some parts of the pipeline were written for the Austrakka platform.
• Phylogenetic trees are constructed using HKY+G
• Ancestral Reconstruction using TreeTime is done using the same settings as dengue.
• The ingest pipeline is designed to fetch only the latest data by checking when the last check was performed. For first time runs, set this date to 1900-01-01 in ingest/resources/ncbi_date_last_checked.txt.
• A nextclade dataset was generated using all high quality jev data as of May 2024. This can be found in nextclade_dataset/resources/nc_dataset.
• Clockrates (in phylogenetics/config/config.yaml) were extracted from this paper:

  gt1: "4.13e-4"
  gt2: "2.41e-4"
  gt3: "6.17e-5"
  gt4: "2.41e-4"
  gt5: "2.41e-4"