Pangenome Guided Genome Assembly Workflow

Directory Structure

• workflow/: Contains the main Snakefile (Other snakefiles are inside other_snakefiles/)
• config/: Contains snakemake configuration files for different datasets
• benchmark/: Runtime and resource usage statistics for individual Snakemake rules
• logs/: Standard output and error logs for each rule execution, useful for debugging
• results*/: Output files generated by the pipeline.

Run the workflow

From the pga_workflow/ directory, run:

snakemake --use-singularity --singularity-args "-B /private/groups/" --configfile config/config_ontR10.yaml --cores 128 --printshellcmds

To resume the workflow from where it left off (e.g., after interruption), use:

snakemake --use-singularity --singularity-args "-B /private/groups/" --configfile config/config_ontR10.yaml --cores 128 --printshellcmds --rerun-incomplete 
## Sometimes, it asks to unlock and then rerun
snakemake --use-singularity --singularity-args "-B /private/groups/" --configfile config/config_ontR10.yaml --cores 128 --printshellcmds --unlock

This will rerun any jobs that were incomplete or failed during the previous run.

Configuration

The workflow is configured using the config/config_ontR10.yaml file. This file contains paths to input files, tool parameters, and other settings. The config file also contains details on the required format of the inputs.

Key Configuration Parameters

• SAMPLE_IDS: A list of sample IDs to process.
• graph_base: Path to the pangenome graph.
• chm13_ref: Path to the CHM13 reference genome. (T2T-CHM13v2.0 currently being used)
• HG002v101_ref: Path to the HG002 reference genome. (HG002v1.0.1 currently being used)
• MINIMAP: Parameters for minimap2.
• HAPLOTYPE_SAMPLING: Parameters for haplotype sampling.
• ANALYSEPAF, DISPLAYPAF, SHASTA: Paths to binaries.
• VG_ANCHORS: Specify path to the config file (default: config/vg_anchors_config.ini)
• TMPDIR: A temporary directory for intermediate files.
• region: Defines the genomic region to be assembled.

Workflow Behavior Flags

Additional flags in the config file allow for fine-grained control over the workflow's behavior:

• RUN_MODE: Controls which workflow to run

  • "no_positive_control": Runs only the sample workflow
  • "positive_control_only": Runs only the positive control workflow. Uses the HG002-included graph for analysis

• READ_TYPE: Controls which type of reads to use.

  • "raw": Use the original, raw sequencing reads.
  • "ec": Use error-corrected reads generated by a tool like hifiasm.

• RUN_GBZ_QUERY (boolean): If True, the workflow will use gbz-base query to extract subgraph files (pg.vg, gfa, dist, gaf). If False, it will use vg chunk.

• USE_FULL_GRAPH (boolean): If True, anchor generation will be performed using the full graph along with the reads and index from the gbz query. Note: This will overwrite the anchors/ directory.

• RUN_DEBUGGING (boolean): If True, the workflow will generate extra output files useful for debugging, such as reliable snarl stats and files for visualization in Bandage.