workflow.md

pipeline overview

Below is a rough overview of what the main part of the pipeline does:

graph TD
    Dw("Download genomes")

    Genomad --> MGEs
    ISEScan --> MGEs
    DefenseFinder --> MGEs
    IntegronFinder --> MGEs

    Dw --> Genomad
    Dw --> ISEScan
    Dw --> DefenseFinder
    Dw --> IntegronFinder

    P --> J("extract junctions")

    J --> JG("build junction graphs")

    JG --> AM("assign MGEs")

    AM --> CS("select binary junctions")

    Dw --> P("build pangraph")

    P --> CGA("extract core genome alignment")

    CGA --> FCGA("filtered core-genome alignment")

    FCGA --> T("tree")

    MGEs --> AM

    CS --> C("infer gain/loss")
    T --> C

Loading

• Genomes are downloaded from NCBI
• Mobile Genetic Elements (MGEs) and defense systems are annotated using Genomad, ISEScan, DefenseFinder and IntegronFinder
• Using pangraph, we build a pangenome graph including all of the chromosomes
• From the graph we extract the core-genome alignment, we filter out highly mutated regions (putative recombination) and build a core-genome tree
• From the graph we extract all junctions, build a graph for each junction
• Junction graphs ar characterized in terms of number of distinct paths, total pangenome content and MGE presence
• We then consider binary junctions and combine the pattern of paths with the core-genome tree to infer gain/loss events

This information is then collected and presented in the main figures produced by the pipeline.

Here you have a full view of the rule graph of snakemake.