Example analysis

This walkthrough demonstrates a typical methylphase run on a mod-tagged BAM with motif-driven typing.

Inputs

• sample.mod.bam + index
• Motifs file motifs.tsv (columns: motif, mod_type, mod_position, optional complement, bin, reference)
• Floria haploset haplotypes.hapset

Steps

1) (Optional) Inspect contigs

   methylphase utils contigs --bam sample.mod.bam

2) (Optional) Extract per-read methylation

   methylphase extract \
     --bam sample.mod.bam \
     --motif-file motifs.tsv \
     --output-dir extract \
     --motif-summary

3) Run the full phase-variants pipeline

   methylphase phase-variants \
     --floria floria/contig/contig.haplosets \
     --bam sample.mod.bam \
     --motif-file motifs.tsv \
     --out phase_variants 

4) Review clustering and model choice

• phase_variants/split_reads/read_clustering_raw.tsv for per-read motif features and clusters.
• phase_variants/fits/ for per-class log-likelihoods and criteria.
• phase_variants/summary.json for the selected model and criteria values.

5) Use the imputed outputs

• imputed.tsv for completed datasets (continuous and categorical fields).
• imputed_labels.tsv for MAP class labels.
• Feed these into downstream association tests or visualization tools.

6) Split FASTQs by inferred class for downstream assembly/analysis of each variant: bash methylphase typing split-fastq \ --fastq reads.fastq \ --responsibilities phase_variants/best_responsibilities.tsv \ --output-dir phase_variants/fastq_by_class