Nextflow entry point: build + persist a joined assembly (file outputs only)

Reads the per-scaffold assembly FASTA + coverage CSVs + the reference FASTA, joins, and writes the joined `ID_assembly_0.fasta`, `ID_assembly_0_coverageStats.csv`, and a `ID_joined_row.csv` (one row of `assemblies` columns) for the workflow to `sqlInsert`. No DB connection is opened here; DB writes are done in Groovy (see scaffold_join_workflow.nf).

Usage

run_scaffold_join(
  assembly_fasta,
  coverage_csvs,
  ref_fasta,
  ID,
  out_dir,
  gap_len = 100,
  circular = FALSE,
  db = NULL,
  auto_join = TRUE,
  scaffold_hits = NULL
)

Arguments

assembly_fasta

path to the multi-scaffold assembly FASTA (headers `ID.path.scaffold [topology]`).

coverage_csvs

character vector of per-scaffold coverageStats CSV paths.

ref_fasta

path to the reference mitogenome FASTA.

ID

sample id.

out_dir

directory to write outputs into.

gap_len

default N-gap length for unknown junctions.

circular

whether to mark the joined molecule circular.

db

optional DBI connection; unused for file-only outputs (DB writes are done in Groovy) and kept for call-site compatibility.

auto_join

when TRUE (and the scaffolds' BLAST hits agree) build + write the joined Path 0. When FALSE (toggle off) or the hits disagree, only the precomputed scaffold->reference mappings are written, leaving the sample fragmented for manual review in the app.

scaffold_hits

optional ';'-separated string of per-scaffold BLAST hits ("scaffold|accession|pident") passed from the workflow. Preferred over a DB read of `assemblies.blast_accession`, which is written by an async UPDATE with no happens-before relative to this step (so it can still be NULL here).