diff --git a/DESCRIPTION b/DESCRIPTION
index ee2f1c5..e7e92a2 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -56,4 +56,4 @@ Suggests:
     rmarkdown,
     writexl,
     testthat (>= 3.0.0)
-Config/roxygen2/version: 8.0.0
+Config/roxygen2/version: 8.1.0
diff --git a/NAMESPACE b/NAMESPACE
index cd61148..aedaaa0 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -4,6 +4,7 @@ export(.extractAMRtable)
 export(.updateBVBRCdata)
 export(CDHIT2duckdb)
 export(buildClusterFeatureMap)
+export(buildDyadFeatureMap)
 export(cleanData)
 export(cleanMetaData)
 export(exportProcessedData)
diff --git a/R/data_processing.R b/R/data_processing.R
index 12df346..45875c2 100644
--- a/R/data_processing.R
+++ b/R/data_processing.R
@@ -970,7 +970,6 @@ CDHIT2duckdb <- function(duckdb_path,
 #' @returns A list of paths to the database hmm files.
 #'
 #' @keywords internal
-#' @examples
 .prepareHmmerDatabases <- function(
     hmmer_db_dir,
     databases = c("Pfam", "COG", "AMRFinder"),
@@ -1281,17 +1280,17 @@ CDHIT2duckdb <- function(duckdb_path,
 
 #' The function to run HMMER with docker
 #'
-#' @param JOB_NAME
-#' @param FASTA
-#' @param DB
-#' @param Total_proteins
-#' @param output_path
-#' @param db_paths
-#' @param docker_image
-#' @param threads
-#' @param n_workers
+#' @param JOB_NAME protein_chunk id
+#' @param FASTA fasta sequences
+#' @param DB HMM database
+#' @param Total_proteins protein sequence count
+#' @param output_path path for saving hmmer outputs
+#' @param db_paths path to HMM database 
+#' @param docker_image hmmer docker image (ideally from dockerhub)
+#' @param threads number of threads
+#' @param n_workers number of parallel workers
 #'
-#' @returns
+#' @returns the filename of the parquet file with hmmer output post parsing
 #'
 #' @keywords internal
 .runHmmerJob <- function(JOB_NAME, FASTA, DB, Total_proteins,
@@ -1375,19 +1374,16 @@ CDHIT2duckdb <- function(duckdb_path,
 
 #' Wrapper for preparing HMM databases and running HMMER on protein sequences from duckdb and writing them.
 #'
-#' @param duckdb_path
-#' @param output_path
-#' @param threads
-#' @param hmmer_db_dir
-#' @param databases
-#' @param docker_image
-#' @param num_of_splits
-#' @param n_workers
-#'
-#' @returns
+#' @param duckdb_path path to the duckdb with protein sequences and list
+#' @param output_path path where HMMER output will be saved
+#' @param threads number of threads
+#' @param hmmer_db_dir path to the directory where HMM databases are/will be downloaded
+#' @param databases list of HMM databases
+#' @param docker_image the docker image of HMMER
+#' @param num_of_splits The number of splits of the protein sequence file for parallel processing
+#' @param n_workers The number of parallel runs
 #'
 #' @keywords internal
-#' @examples
 .runHMMER <- function(duckdb_path,
                       output_path,
                       threads = 8L,
@@ -2203,16 +2199,18 @@ CDHIT2duckdb <- function(duckdb_path,
 }
 
 
-# Clean BV-BRC metadata, then save as Parquet files
-#'
-#' @param duckdb_path
-#' @param path
-#' @param ref_file_path
+#' Clean BV-BRC metadata, then save as Parquet files
 #'
-#' @returns
+#' @param duckdb_path Path to the **per-selection DuckDB** produced by
+#'   [prepareGenomes()] (e.g., `"data/<Bug>/<Abbrev>.duckdb"`). This DB must
+#'   already contain the tables written by [prepareGenomes()] and the upstream
+#'   genome-processing steps.
+#' @param path the path to working directory
+#' @param ref_file_path Directory containing reference TSVs used by
+#'   [cleanMetaData()] and [cleanData()] for metadata harmonization.
+#'   Default: `"data_raw/"`.
+#' 
 #' @export
-#'
-#' @examples
 cleanMetaData <- function(duckdb_path, path, ref_file_path = "data_raw/") {
   duckdb_path <- normalizePath(duckdb_path)
   # If no explicit path is provided (or a generic one), choose results/<bug>/ when
@@ -2358,15 +2356,15 @@ cleanMetaData <- function(duckdb_path, path, ref_file_path = "data_raw/") {
   invisible(TRUE)
 }
 
-# Clean feature matrices, then save as Parquet files
+#' Clean feature matrices, then save as Parquet files
 #'
-#' @param duckdb_path
-#' @param path
+#' @param duckdb_path Path to the **per-selection DuckDB** produced by
+#'   [prepareGenomes()] (e.g., `"data/<Bug>/<Abbrev>.duckdb"`). This DB must
+#'   already contain the tables written by [prepareGenomes()] and the upstream
+#'   genome-processing steps.
+#' @param path the path to working directory
 #'
-#' @returns
 #' @export
-#'
-#' @examples
 cleanData <- function(duckdb_path, path) {
   duckdb_path <- normalizePath(duckdb_path)
   # If no explicit path is provided (or a generic one), choose results/<bug>/ when
@@ -2513,6 +2511,7 @@ cleanData <- function(duckdb_path, path) {
         names_to = "annotation",
         values_to = "value"
       ) |>
+      dplyr::rename(!!database := annotation) |>
       dplyr::filter(!is.na(value) & value != "") |>
       dplyr::mutate(value = as.integer(value)) |>
       writeCompressedParquet(count_parquet)
diff --git a/R/feature_to_head.R b/R/feature_to_head.R
new file mode 100644
index 0000000..711d796
--- /dev/null
+++ b/R/feature_to_head.R
@@ -0,0 +1,612 @@
+#' Build a protein-gene dyad feature network using DuckDB
+#'
+#' Constructs a two-mode (bipartite) network in which every node is either a
+#' protein-gene dyad or a biological feature, and every edge links a dyad to one
+#' of its features. Features include the dyad's own protein and gene, structural
+#' (pangenome graph) gene arrangements, and HMMER annotations: protein domains
+#' (Pfam), COGs, phage-defense systems (including Cas), and antimicrobial
+#' resistance genes.
+#'
+#' @param duckdb_path Character. Path to the source per-selection DuckDB. The
+#'   associated Parquet files and provenance manifest are expected to live
+#'   alongside it. This is the same `duckdb_path` produced by
+#'   \code{\link{runDataProcessing}}.
+#' @param additional_feature_scales Character vector of optional feature types to
+#'   include. If `NULL`, all HMMER databases recorded in the manifest are used,
+#'   plus `struct` to include pangenome graph triplets. If a character vector is
+#'   supplied, individual feature scales can be selected.
+#' @param output_path Character or NULL. Directory where the output Parquet file
+#'   will be written. If NULL, the output is written alongside the source DuckDB.
+#'
+#' @details
+#' The function performs the following steps:
+#' \enumerate{
+#'   \item Creates a protein-gene mapping from
+#'     \code{genome_gene_protein.parquet}.
+#'   \item Constructs a unique protein-gene dyad identifier of the form
+#'     \code{"protein|gene"}.
+#'   \item Optionally loads structural and HMMER annotations.
+#'   \item Generates a network edge list linking each dyad to its associated
+#'     features.
+#'   \item Writes the resulting edge list to a compressed Parquet file.
+#' }
+#'
+#' Every edge runs from a `protein|gene` dyad to a type-prefixed feature node:
+#'
+#' \preformatted{
+#' protein|gene --> protein:PROTEIN_ID
+#' protein|gene --> gene:GENE_ID
+#' protein|gene --> pfam:PFXXXXX
+#' protein|gene --> cog:COGXXXX
+#' protein|gene --> amr:GENE_NAME
+#' protein|gene --> defensecas:DEFENSE_SYSTEM
+#' protein|gene --> struct:STRUCTURE
+#' }
+#'
+#' The type prefix on each target keeps feature namespaces separate, so a
+#' generic feature name cannot collide across scales downstream.
+#'
+#' The output edge list contains two columns:
+#' \describe{
+#'   \item{source}{Protein-gene dyad identifier.}
+#'   \item{target}{Feature node identifier prefixed by feature type.}
+#' }
+#'
+#' @return Invisibly returns the path to the generated
+#'   \code{dyad_feature.parquet} file.
+#'
+#' @examples
+#' \dontrun{
+#' buildDyadFeatureMap(
+#'   duckdb_path = "data/Staphylococcus_argenteus/Sar.duckdb"
+#' )
+#' }
+#'
+#' @import DBI duckdb
+#' @export
+buildDyadFeatureMap <- function(
+    duckdb_path,
+    additional_feature_scales = NULL,
+    output_path = NULL
+) {
+
+  duckdb_path <- normalizePath(
+    duckdb_path,
+    mustWork = TRUE
+  )
+
+  parquet_dir <- dirname(duckdb_path)
+
+  manifest_path <- .manifest_find_latest(
+    duckdb_path
+  )
+
+  if (is.null(manifest_path)) {
+    stop(
+      "No provenance manifest found for: ",
+      duckdb_path
+    )
+  }
+
+  # Read in the `runDataProcessing()` output manifest
+  manifest <- jsonlite::read_json(
+    manifest_path,
+    simplifyVector = FALSE
+  )
+
+  hmmer_stage <- NULL
+
+  # Find the most recent run whose HMMER stage completed successfully; its
+  # recorded databases determine which feature scales are available.
+  for (run in rev(manifest$runs)) {
+
+    stages <- run$stages %||% list()
+
+    matches <- purrr::keep(
+      stages,
+      ~ identical(.x$name, "hmmer") &&
+        identical(.x$status, "success")
+    )
+
+    if (length(matches)) {
+      hmmer_stage <- matches[[1]]
+      break
+    }
+  }
+
+  # Without a successful HMMER stage the annotation Parquets we join against
+  # are not guaranteed to exist, so there is nothing to build from.
+  if (is.null(hmmer_stage)) {
+    stop(
+      "No successful HMMER stage found in manifest: ",
+      manifest_path
+    )
+  }
+
+  hmmer_databases <- unlist(
+    hmmer_stage$parameters$databases
+  )
+
+  # A successful stage with no databases recorded should not happen; fail loudly.
+  if (!length(hmmer_databases)) {
+    stop(
+      "HMMER stage in manifest does not contain any databases."
+    )
+  }
+
+  # Feature scales the caller is allowed to request: the structural view plus
+  # every HMMER database from the manifest (so new databases are picked up
+  # automatically).
+  allowed_features <- c(
+    "struct",
+    hmmer_databases
+  )
+
+  # Default to every allowed scale; otherwise reject anything unrecognised.
+  if (is.null(additional_feature_scales)) {
+    additional_feature_scales <- allowed_features
+  } else {
+    unknown_features <- setdiff(
+      additional_feature_scales,
+      allowed_features
+    )
+
+    if (length(unknown_features)) {
+      stop(
+        "Unsupported feature scale(s): ",
+        paste(unknown_features, collapse = ", "),
+        ". Available features: ",
+        paste(allowed_features, collapse = ", ")
+      )
+    }
+  }
+
+  out_dir <- if (is.null(output_path)) {
+    parquet_dir
+  } else {
+    normalizePath(
+      output_path,
+      winslash = "/",
+      mustWork = FALSE
+    )
+  }
+
+  dir.create(
+    out_dir,
+    recursive = TRUE,
+    showWarnings = FALSE
+  )
+
+  parquet_path <- file.path(
+    out_dir,
+    "dyad_feature.parquet"
+  )
+
+  con <- DBI::dbConnect(
+    duckdb::duckdb(),
+    dbdir = ":memory:"
+  )
+
+  on.exit(
+    DBI::dbDisconnect(
+      con,
+      shutdown = TRUE
+    ),
+    add = TRUE
+  )
+
+  # Local helpers -------------------------------------------------------------
+
+  # Escape single quotes so a path can be embedded in a SQL string literal.
+  .sql_escape <- function(x) {
+    gsub(
+      "'",
+      "''",
+      x,
+      fixed = TRUE
+    )
+  }
+
+  # Resolve "<parquet_dir>/<dataset_name>.parquet" to an escaped absolute path,
+  # erroring if the expected Parquet file is missing.
+  .parquet_dataset_sql <- function(
+    parquet_dir,
+    dataset_name
+  ) {
+
+    path <- file.path(
+      parquet_dir,
+      paste0(
+        dataset_name,
+        ".parquet"
+      )
+    )
+
+    if (!file.exists(path)) {
+      stop(
+        "Required Parquet file not found: ",
+        path
+      )
+    }
+
+    .sql_escape(
+      normalizePath(
+        path,
+        winslash = "/",
+        mustWork = TRUE
+      )
+    )
+  }
+
+  # =========================
+  # Gene -> protein mapping
+  # =========================
+
+  # Base protein/gene pairs. Empty strings are excluded alongside NULLs to match
+  # the `value != ""` convention in data_processing.R and to avoid emitting
+  # bogus "protein|" or "|gene" dyads.
+  sql_path <- .parquet_dataset_sql(
+    parquet_dir,
+    "genome_gene_protein"
+  )
+
+  DBI::dbExecute(
+    con,
+    sprintf(
+      "
+      CREATE OR REPLACE VIEW protein_gene AS
+      SELECT DISTINCT
+        protein_ids AS protein,
+        REPLACE(Gene, '~', '.') AS gene
+      FROM read_parquet('%s')
+      WHERE protein_ids IS NOT NULL
+        AND protein_ids <> ''
+        AND Gene IS NOT NULL
+        AND Gene <> ''
+      ",
+      sql_path
+    )
+  )
+
+  # =========================
+  # Protein-gene -> dyad
+  # =========================
+
+  # Collapse each protein/gene pair into a single "protein|gene" dyad id; this
+  # is the source node for every edge in the output network.
+  DBI::dbExecute(
+    con,
+    "
+    CREATE OR REPLACE VIEW protein_gene_dyad AS
+    SELECT DISTINCT
+      protein,
+      gene,
+      CONCAT(protein, '|', gene) AS dyad
+    FROM protein_gene
+    "
+  )
+
+  # Track which optional feature views were successfully created, so the edge
+  # queries below only join against views that exist.
+  has_struct <- FALSE
+  has_pfam <- FALSE
+  has_cog <- FALSE
+  has_amr <- FALSE
+  has_defensecas <- FALSE
+
+  # =========================
+  # Structural gene features
+  # =========================
+
+  if ("struct" %in% additional_feature_scales) {
+
+    struct_path <- file.path(
+      parquet_dir,
+      "struct.parquet"
+    )
+
+    if (!file.exists(struct_path)) {
+
+      # Same skip-and-continue behaviour as .create_feature_view() uses for the
+      # HMMER datasets, kept inline here because the struct view is built with a
+      # different (UNNEST) query.
+      message(
+        "Skipping struct: no parquet found. Generate struct parquet first."
+      )
+
+    } else {
+
+      sql_path <- .parquet_dataset_sql(
+        parquet_dir,
+        "struct"
+      )
+
+      DBI::dbExecute(
+        con,
+        sprintf(
+          "
+          CREATE OR REPLACE VIEW v_struct_genes AS
+          SELECT DISTINCT
+            struct,
+            gene
+          FROM read_parquet('%s') s
+          CROSS JOIN UNNEST(
+            string_split(s.struct, '.')
+          ) AS t(gene)
+          WHERE s.value = 1
+          ",
+          sql_path
+        )
+      )
+
+      has_struct <- TRUE
+    }
+  }
+
+  # ==========================
+  # Generic HMMER feature view
+  # ==========================
+
+  # Local helper: build a `protein -> feature` view from one HMMER annotation
+  # Parquet. `feature_expr` is the SQL expression that derives the feature label
+  # from `query_name` (identity for most databases, a REPLACE() for AMRFinder).
+  # Returns TRUE if the view was created, FALSE if the Parquet was missing.
+  .create_feature_view <- function(
+    con,
+    view_name,
+    parquet_dir,
+    dataset_name,
+    feature_expr
+  ) {
+
+    sql_path <- tryCatch(
+      .parquet_dataset_sql(
+        parquet_dir,
+        dataset_name
+      ),
+      error = function(e) {
+        message(
+          "Skipping ",
+          dataset_name,
+          ": ",
+          e$message
+        )
+        NULL
+      }
+    )
+
+    if (is.null(sql_path)) {
+      return(FALSE)
+    }
+
+    DBI::dbExecute(
+      con,
+      sprintf(
+        "
+        CREATE OR REPLACE VIEW %s AS
+        SELECT DISTINCT
+          protein,
+          %s AS feature
+        FROM read_parquet('%s')
+        WHERE protein IS NOT NULL
+          AND query_name IS NOT NULL
+        ",
+        view_name,
+        feature_expr,
+        sql_path
+      )
+    )
+
+    TRUE
+  }
+
+  # =========================
+  # HMMER annotation features
+  # =========================
+
+  if ("Pfam" %in% additional_feature_scales) {
+
+    has_pfam <- .create_feature_view(
+      con = con,
+      view_name = "v_pfam",
+      parquet_dir = parquet_dir,
+      dataset_name = "protein_Pfam",
+      feature_expr = "query_name"
+    )
+  }
+
+  if ("COG" %in% additional_feature_scales) {
+
+    has_cog <- .create_feature_view(
+      con = con,
+      view_name = "v_cog",
+      parquet_dir = parquet_dir,
+      dataset_name = "protein_COG",
+      feature_expr = "query_name"
+    )
+  }
+
+  if ("AMRFinder" %in% additional_feature_scales) {
+
+    has_amr <- .create_feature_view(
+      con = con,
+      view_name = "v_amrfinder",
+      parquet_dir = parquet_dir,
+      dataset_name = "protein_AMRFinder",
+      feature_expr = "REPLACE(REPLACE(query_name, '-NCBIFAM', ''), '-', '.')"
+    )
+  }
+
+  if ("DefenseCas" %in% additional_feature_scales) {
+
+    has_defensecas <- .create_feature_view(
+      con = con,
+      view_name = "v_defensecas",
+      parquet_dir = parquet_dir,
+      dataset_name = "protein_DefenseCas",
+      feature_expr = "query_name"
+    )
+  }
+
+  # =========================
+  # Build network edge queries
+  # =========================
+
+  # Each entry is a SELECT returning (source, target) rows that are UNIONed into
+  # the final edge list. In the joined queries below `pgd` aliases the
+  # `protein_gene_dyad` view, so `pgd.dyad` is the "protein|gene" dyad id.
+  edge_queries <- c(
+    "
+    SELECT DISTINCT
+      dyad AS source,
+      CONCAT('protein:', protein) AS target
+    FROM protein_gene_dyad
+    ",
+    "
+    SELECT DISTINCT
+      dyad AS source,
+      CONCAT('gene:', gene) AS target
+    FROM protein_gene_dyad
+    "
+  )
+
+  # =========================
+  # Structure edges
+  # =========================
+
+  if (has_struct) {
+
+    edge_queries <- c(
+      edge_queries,
+      "
+      SELECT DISTINCT
+        pgd.dyad AS source,
+        CONCAT('struct:', sg.struct) AS target
+      FROM protein_gene_dyad pgd
+      JOIN v_struct_genes sg
+        ON pgd.gene = sg.gene
+      "
+    )
+  }
+
+  # =========================
+  # Pfam edges
+  # =========================
+
+  if (has_pfam) {
+
+    edge_queries <- c(
+      edge_queries,
+      "
+      SELECT DISTINCT
+        pgd.dyad AS source,
+        CONCAT('pfam:', pf.feature) AS target
+      FROM protein_gene_dyad pgd
+      JOIN v_pfam pf
+        ON pgd.protein = pf.protein
+      "
+    )
+  }
+
+  # =========================
+  # COG edges
+  # =========================
+
+  if (has_cog) {
+
+    edge_queries <- c(
+      edge_queries,
+      "
+      SELECT DISTINCT
+        pgd.dyad AS source,
+        CONCAT('cog:', cf.feature) AS target
+      FROM protein_gene_dyad pgd
+      JOIN v_cog cf
+        ON pgd.protein = cf.protein
+      "
+    )
+  }
+
+  # =========================
+  # AMRFinder edges
+  # =========================
+
+  if (has_amr) {
+
+    edge_queries <- c(
+      edge_queries,
+      "
+      SELECT DISTINCT
+        pgd.dyad AS source,
+        CONCAT('amr:', af.feature) AS target
+      FROM protein_gene_dyad pgd
+      JOIN v_amrfinder af
+        ON pgd.protein = af.protein
+      "
+    )
+  }
+
+  # =========================
+  # DefenseCas edges
+  # =========================
+
+  if (has_defensecas) {
+
+    edge_queries <- c(
+      edge_queries,
+      "
+      SELECT DISTINCT
+        pgd.dyad AS source,
+        CONCAT('defensecas:', df.feature) AS target
+      FROM protein_gene_dyad pgd
+      JOIN v_defensecas df
+        ON pgd.protein = df.protein
+      "
+    )
+  }
+
+  # =========================
+  # Final edge list
+  # =========================
+
+  DBI::dbExecute(
+    con,
+    paste0(
+      "
+      CREATE OR REPLACE VIEW network_edges AS
+      ",
+      paste(
+        edge_queries,
+        collapse = "\nUNION\n"
+      )
+    )
+  )
+
+  # =========================
+  # Export Parquet
+  # =========================
+
+  parquet_sql <- DBI::dbQuoteString(
+    con,
+    normalizePath(
+      parquet_path,
+      winslash = "/",
+      mustWork = FALSE
+    )
+  )
+
+  DBI::dbExecute(
+    con,
+    paste0(
+      "
+      COPY network_edges
+      TO ",
+      parquet_sql,
+      "
+      (FORMAT PARQUET, COMPRESSION ZSTD)
+      "
+    )
+  )
+
+  invisible(parquet_path)
+}
diff --git a/data/Staphylococcus_epidermidis/dyad_feature.parquet b/data/Staphylococcus_epidermidis/dyad_feature.parquet
new file mode 100644
index 0000000..f7951e1
Binary files /dev/null and b/data/Staphylococcus_epidermidis/dyad_feature.parquet differ
diff --git a/data/Staphylococcus_epidermidis/genome_gene_protein.parquet b/data/Staphylococcus_epidermidis/genome_gene_protein.parquet
new file mode 100644
index 0000000..0873eed
Binary files /dev/null and b/data/Staphylococcus_epidermidis/genome_gene_protein.parquet differ
diff --git a/data/Staphylococcus_epidermidis/protein_AMRFinder.parquet b/data/Staphylococcus_epidermidis/protein_AMRFinder.parquet
new file mode 100644
index 0000000..d7bcd87
Binary files /dev/null and b/data/Staphylococcus_epidermidis/protein_AMRFinder.parquet differ
diff --git a/data/Staphylococcus_epidermidis/protein_COG.parquet b/data/Staphylococcus_epidermidis/protein_COG.parquet
new file mode 100644
index 0000000..2350f5d
Binary files /dev/null and b/data/Staphylococcus_epidermidis/protein_COG.parquet differ
diff --git a/data/Staphylococcus_epidermidis/struct.parquet b/data/Staphylococcus_epidermidis/struct.parquet
new file mode 100644
index 0000000..129fa72
Binary files /dev/null and b/data/Staphylococcus_epidermidis/struct.parquet differ
diff --git a/man/buildDyadFeatureMap.Rd b/man/buildDyadFeatureMap.Rd
new file mode 100644
index 0000000..43fcd11
--- /dev/null
+++ b/man/buildDyadFeatureMap.Rd
@@ -0,0 +1,80 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/feature_to_head.R
+\name{buildDyadFeatureMap}
+\alias{buildDyadFeatureMap}
+\title{Build a protein-gene dyad feature network using DuckDB}
+\usage{
+buildDyadFeatureMap(
+  duckdb_path,
+  additional_feature_scales = NULL,
+  output_path = NULL
+)
+}
+\arguments{
+\item{duckdb_path}{Character. Path to the source per-selection DuckDB. The
+associated Parquet files and provenance manifest are expected to live
+alongside it. This is the same \code{duckdb_path} produced by
+\code{\link{runDataProcessing}}.}
+
+\item{additional_feature_scales}{Character vector of optional feature types to
+include. If \code{NULL}, all HMMER databases recorded in the manifest are used,
+plus \code{struct} to include pangenome graph triplets. If a character vector is
+supplied, individual feature scales can be selected.}
+
+\item{output_path}{Character or NULL. Directory where the output Parquet file
+will be written. If NULL, the output is written alongside the source DuckDB.}
+}
+\value{
+Invisibly returns the path to the generated
+\code{dyad_feature.parquet} file.
+}
+\description{
+Constructs a two-mode (bipartite) network in which every node is either a
+protein-gene dyad or a biological feature, and every edge links a dyad to one
+of its features. Features include the dyad's own protein and gene, structural
+(pangenome graph) gene arrangements, and HMMER annotations: protein domains
+(Pfam), COGs, phage-defense systems (including Cas), and antimicrobial
+resistance genes.
+}
+\details{
+The function performs the following steps:
+\enumerate{
+\item Creates a protein-gene mapping from
+\code{genome_gene_protein.parquet}.
+\item Constructs a unique protein-gene dyad identifier of the form
+\code{"protein|gene"}.
+\item Optionally loads structural and HMMER annotations.
+\item Generates a network edge list linking each dyad to its associated
+features.
+\item Writes the resulting edge list to a compressed Parquet file.
+}
+
+Every edge runs from a \code{protein|gene} dyad to a type-prefixed feature node:
+
+\preformatted{
+protein|gene --> protein:PROTEIN_ID
+protein|gene --> gene:GENE_ID
+protein|gene --> pfam:PFXXXXX
+protein|gene --> cog:COGXXXX
+protein|gene --> amr:GENE_NAME
+protein|gene --> defensecas:DEFENSE_SYSTEM
+protein|gene --> struct:STRUCTURE
+}
+
+The type prefix on each target keeps feature namespaces separate, so a
+generic feature name cannot collide across scales downstream.
+
+The output edge list contains two columns:
+\describe{
+\item{source}{Protein-gene dyad identifier.}
+\item{target}{Feature node identifier prefixed by feature type.}
+}
+}
+\examples{
+\dontrun{
+buildDyadFeatureMap(
+  duckdb_path = "data/Staphylococcus_argenteus/Sar.duckdb"
+)
+}
+
+}
diff --git a/man/cleanData.Rd b/man/cleanData.Rd
new file mode 100644
index 0000000..ee2ba2b
--- /dev/null
+++ b/man/cleanData.Rd
@@ -0,0 +1,19 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/data_processing.R
+\name{cleanData}
+\alias{cleanData}
+\title{Clean feature matrices, then save as Parquet files}
+\usage{
+cleanData(duckdb_path, path)
+}
+\arguments{
+\item{duckdb_path}{Path to the \strong{per-selection DuckDB} produced by
+\code{\link[=prepareGenomes]{prepareGenomes()}} (e.g., \code{"data/<Bug>/<Abbrev>.duckdb"}). This DB must
+already contain the tables written by \code{\link[=prepareGenomes]{prepareGenomes()}} and the upstream
+genome-processing steps.}
+
+\item{path}{the path to working directory}
+}
+\description{
+Clean feature matrices, then save as Parquet files
+}
diff --git a/man/cleanMetaData.Rd b/man/cleanMetaData.Rd
new file mode 100644
index 0000000..b2a0912
--- /dev/null
+++ b/man/cleanMetaData.Rd
@@ -0,0 +1,23 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/data_processing.R
+\name{cleanMetaData}
+\alias{cleanMetaData}
+\title{Clean BV-BRC metadata, then save as Parquet files}
+\usage{
+cleanMetaData(duckdb_path, path, ref_file_path = "data_raw/")
+}
+\arguments{
+\item{duckdb_path}{Path to the \strong{per-selection DuckDB} produced by
+\code{\link[=prepareGenomes]{prepareGenomes()}} (e.g., \code{"data/<Bug>/<Abbrev>.duckdb"}). This DB must
+already contain the tables written by \code{\link[=prepareGenomes]{prepareGenomes()}} and the upstream
+genome-processing steps.}
+
+\item{path}{the path to working directory}
+
+\item{ref_file_path}{Directory containing reference TSVs used by
+\code{\link[=cleanMetaData]{cleanMetaData()}} and \code{\link[=cleanData]{cleanData()}} for metadata harmonization.
+Default: \code{"data_raw/"}.}
+}
+\description{
+Clean BV-BRC metadata, then save as Parquet files
+}
diff --git a/man/domainFromIPR.Rd b/man/domainFromIPR.Rd
deleted file mode 100644
index 6b3229a..0000000
--- a/man/domainFromIPR.Rd
+++ /dev/null
@@ -1,23 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_processing.R
-\name{domainFromIPR}
-\alias{domainFromIPR}
-\title{Derive protein domain presence/absence and counts via InterProScan and write to DuckDB}
-\usage{
-domainFromIPR(
-  duckdb_path,
-  path,
-  out_file_base = "iprscan",
-  appl = c("Pfam"),
-  ipr_version = "5.76-107.0",
-  ipr_dest_dir = "inst/extdata/interpro",
-  ipr_platform = "linux/amd64",
-  auto_prepare_data = TRUE,
-  threads = 8,
-  file_format = "TSV",
-  docker_repo = "interpro/interproscan"
-)
-}
-\description{
-Derive protein domain presence/absence and counts via InterProScan and write to DuckDB
-}
diff --git a/man/dot-checkInterProData.Rd b/man/dot-checkInterProData.Rd
deleted file mode 100644
index 7f4c794..0000000
--- a/man/dot-checkInterProData.Rd
+++ /dev/null
@@ -1,36 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_processing.R
-\name{.checkInterProData}
-\alias{.checkInterProData}
-\title{Check or install InterProScan data bundle}
-\usage{
-.checkInterProData(
-  version = "5.76-107.0",
-  dest_dir = "inst/extdata/interpro",
-  docker_image = sprintf("interpro/interproscan:\%s", version),
-  platform = "linux/amd64",
-  curl_bin = "curl",
-  verbose = TRUE
-)
-}
-\arguments{
-\item{version}{InterProScan version string.}
-
-\item{dest_dir}{Directory where data should be installed.}
-
-\item{docker_image}{Docker image string for InterProScan.}
-
-\item{platform}{Character indicating Docker platform (e.g. \code{"linux/amd64"}).}
-
-\item{curl_bin}{Path to curl executable.}
-
-\item{verbose}{Logical; print status messages.}
-}
-\value{
-A list containing \code{data_dir} and \code{ready} status.
-}
-\description{
-Ensures that the InterProScan data directory exists locally, downloading
-and verifying the appropriate tarball when necessary.
-}
-\keyword{internal}
diff --git a/man/dot-defenseHMMER.Rd b/man/dot-defenseHMMER.Rd
index 2a3a772..c2ab6d9 100644
--- a/man/dot-defenseHMMER.Rd
+++ b/man/dot-defenseHMMER.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/data_processing.R
 \name{.defenseHMMER}
 \alias{.defenseHMMER}
 \title{Annotate proteins using DefenseFinder + CasFinder HMMs
diff --git a/man/dot-docker_path.Rd b/man/dot-docker_path.Rd
index d0b6f70..c3385ca 100644
--- a/man/dot-docker_path.Rd
+++ b/man/dot-docker_path.Rd
@@ -1,21 +1,12 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_curation.R, R/data_processing.R
+% Please edit documentation in R/helpers.R
 \name{.docker_path}
 \alias{.docker_path}
 \title{Helps normalize Docker paths}
 \usage{
 .docker_path(p)
-
-.docker_path(p)
-}
-\arguments{
-\item{p}{Character scalar. A filesystem path on the host OS.}
-}
-\value{
-A normalized path string.
 }
 \description{
-Converts Windows and mixed-separator paths to forward slashes
-and applies \code{normalizePath()} without requiring the path to exist.
+Helps normalize Docker paths
 }
 \keyword{internal}
diff --git a/man/dot-getDfIPRColNames.Rd b/man/dot-getDfIPRColNames.Rd
deleted file mode 100644
index 7151d2d..0000000
--- a/man/dot-getDfIPRColNames.Rd
+++ /dev/null
@@ -1,19 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_processing.R
-\name{.getDfIPRColNames}
-\alias{.getDfIPRColNames}
-\title{Internal helpers for reading InterProScan TSV outputs}
-\usage{
-.getDfIPRColNames()
-}
-\arguments{
-\item{filepath}{Path to a \code{.tsv} or \code{.tsv.gz} InterProScan result file.}
-}
-\value{
-A tibble of parsed InterProScan output.
-}
-\description{
-Provide standardized column names, types, and a reader wrapper for the
-InterProScan tab-delimited output format.
-}
-\keyword{internal}
diff --git a/man/dot-getDfIPRColTypes.Rd b/man/dot-getDfIPRColTypes.Rd
deleted file mode 100644
index c827bc1..0000000
--- a/man/dot-getDfIPRColTypes.Rd
+++ /dev/null
@@ -1,19 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_processing.R
-\name{.getDfIPRColTypes}
-\alias{.getDfIPRColTypes}
-\title{Internal helpers for reading InterProScan TSV outputs}
-\usage{
-.getDfIPRColTypes()
-}
-\arguments{
-\item{filepath}{Path to a \code{.tsv} or \code{.tsv.gz} InterProScan result file.}
-}
-\value{
-A tibble of parsed InterProScan output.
-}
-\description{
-Provide standardized column names, types, and a reader wrapper for the
-InterProScan tab-delimited output format.
-}
-\keyword{internal}
diff --git a/man/dot-id_checker.Rd b/man/dot-id_checker.Rd
index 00fd8eb..aa4a11e 100644
--- a/man/dot-id_checker.Rd
+++ b/man/dot-id_checker.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_curation.R
+% Please edit documentation in R/helpers.R
 \name{.id_checker}
 \alias{.id_checker}
 \title{Helps ensure trailing 0s are retained in genome IDs for proper downloading}
diff --git a/man/dot-isValidHmmFile.Rd b/man/dot-isValidHmmFile.Rd
index 1268a9b..4ba8824 100644
--- a/man/dot-isValidHmmFile.Rd
+++ b/man/dot-isValidHmmFile.Rd
@@ -1,12 +1,19 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/helpers.R
 \name{.isValidHmmFile}
 \alias{.isValidHmmFile}
-\title{Validate if a HMM file has old HMMER3 format and remove them}
+\title{Validate if a HMM file has old HMMER3 format}
 \usage{
 .isValidHmmFile(hmm_file)
 }
+\arguments{
+\item{hmm_file}{Path to a \code{.hmm} file.}
+}
+\value{
+\code{TRUE} if the file has valid HMMER3 formatting (starts with
+\code{HMMER3/f} and ends with \verb{//}), \code{FALSE} otherwise.
+}
 \description{
-Validate if a HMM file has old HMMER3 format and remove them
+Validate if a HMM file has old HMMER3 format
 }
 \keyword{internal}
diff --git a/man/dot-manifest_event.Rd b/man/dot-manifest_event.Rd
new file mode 100644
index 0000000..fc2182d
--- /dev/null
+++ b/man/dot-manifest_event.Rd
@@ -0,0 +1,24 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_event}
+\alias{.manifest_event}
+\title{Append a provenance event to the active manifest run}
+\usage{
+.manifest_event(manifest_state, level = "info", message, details = list())
+}
+\arguments{
+\item{manifest_state}{Manifest state returned by \code{\link[=.manifest_start]{.manifest_start()}}.}
+
+\item{level}{Character event level.}
+
+\item{message}{Character message.}
+
+\item{details}{Optional named list.}
+}
+\value{
+Updated manifest state.
+}
+\description{
+Append a provenance event to the active manifest run
+}
+\keyword{internal}
diff --git a/man/dot-manifest_file_info.Rd b/man/dot-manifest_file_info.Rd
new file mode 100644
index 0000000..69e9f77
--- /dev/null
+++ b/man/dot-manifest_file_info.Rd
@@ -0,0 +1,20 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_file_info}
+\alias{.manifest_file_info}
+\title{Returns the basics about a file for manifest logging}
+\usage{
+.manifest_file_info(path, hash = FALSE)
+}
+\arguments{
+\item{path}{Character vector of file paths.}
+
+\item{hash}{Logical. If TRUE, calculate MD5 checksums.}
+}
+\value{
+A list of file records.
+}
+\description{
+Returns the basics about a file for manifest logging
+}
+\keyword{internal}
diff --git a/man/dot-manifest_finish.Rd b/man/dot-manifest_finish.Rd
new file mode 100644
index 0000000..25f6495
--- /dev/null
+++ b/man/dot-manifest_finish.Rd
@@ -0,0 +1,22 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_finish}
+\alias{.manifest_finish}
+\title{Finish an active provenance manifest run}
+\usage{
+.manifest_finish(manifest_state, status = "success", error = NULL)
+}
+\arguments{
+\item{manifest_state}{Manifest state returned by \code{\link[=.manifest_start]{.manifest_start()}}.}
+
+\item{status}{Final run status.}
+
+\item{error}{Optional error message.}
+}
+\value{
+Invisibly returns the final manifest state.
+}
+\description{
+Finish an active provenance manifest run
+}
+\keyword{internal}
diff --git a/man/dot-manifest_git_info.Rd b/man/dot-manifest_git_info.Rd
new file mode 100644
index 0000000..937c627
--- /dev/null
+++ b/man/dot-manifest_git_info.Rd
@@ -0,0 +1,18 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_git_info}
+\alias{.manifest_git_info}
+\title{Capture basic GitHub repo state for manifest provenance}
+\usage{
+.manifest_git_info(base_dir = ".")
+}
+\arguments{
+\item{base_dir}{Character. Project root.}
+}
+\value{
+A named list.
+}
+\description{
+Capture basic GitHub repo state for manifest provenance
+}
+\keyword{internal}
diff --git a/man/dot-manifest_package_versions.Rd b/man/dot-manifest_package_versions.Rd
new file mode 100644
index 0000000..4df2167
--- /dev/null
+++ b/man/dot-manifest_package_versions.Rd
@@ -0,0 +1,15 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_package_versions}
+\alias{.manifest_package_versions}
+\title{Capture package versions currently loaded in the R session}
+\usage{
+.manifest_package_versions()
+}
+\value{
+Named character vector of package versions.
+}
+\description{
+Capture package versions currently loaded in the R session
+}
+\keyword{internal}
diff --git a/man/dot-manifest_resume.Rd b/man/dot-manifest_resume.Rd
new file mode 100644
index 0000000..46f28b5
--- /dev/null
+++ b/man/dot-manifest_resume.Rd
@@ -0,0 +1,22 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_resume}
+\alias{.manifest_resume}
+\title{Resume provenance logging in an existing manifest}
+\usage{
+.manifest_resume(manifest_path, base_dir = ".", hash_files = FALSE)
+}
+\arguments{
+\item{manifest_path}{Character. Path to an existing JSON manifest.}
+
+\item{base_dir}{Character. Project root.}
+
+\item{hash_files}{Logical. Calculate SHA-256 checksums for manifest-recorded files.}
+}
+\value{
+A manifest object with \code{path} and \code{run_index}.
+}
+\description{
+Loads an existing manifest and appends a new run.
+}
+\keyword{internal}
diff --git a/man/dot-manifest_run_id.Rd b/man/dot-manifest_run_id.Rd
new file mode 100644
index 0000000..28d860c
--- /dev/null
+++ b/man/dot-manifest_run_id.Rd
@@ -0,0 +1,15 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_run_id}
+\alias{.manifest_run_id}
+\title{Generate a unique manifest run identifier}
+\usage{
+.manifest_run_id()
+}
+\value{
+Character scalar.
+}
+\description{
+Generate a unique manifest run identifier
+}
+\keyword{internal}
diff --git a/man/dot-manifest_stage.Rd b/man/dot-manifest_stage.Rd
new file mode 100644
index 0000000..803a3fd
--- /dev/null
+++ b/man/dot-manifest_stage.Rd
@@ -0,0 +1,44 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_stage}
+\alias{.manifest_stage}
+\title{Update a manifest stage}
+\usage{
+.manifest_stage(
+  manifest_state,
+  name,
+  status = "success",
+  parameters = list(),
+  inputs = character(),
+  outputs = character(),
+  tool = list(),
+  metrics = list(),
+  message = NULL
+)
+}
+\arguments{
+\item{manifest_state}{Manifest state returned by \code{\link[=.manifest_start]{.manifest_start()}}.}
+
+\item{name}{Character stage name.}
+
+\item{status}{Character stage status.}
+
+\item{parameters}{Optional named list.}
+
+\item{inputs}{Optional character vector of input paths.}
+
+\item{outputs}{Optional character vector of output paths.}
+
+\item{tool}{Optional named list describing the tool.}
+
+\item{metrics}{Optional named list of metrics.}
+
+\item{message}{Optional log message.}
+}
+\value{
+Updated manifest state.
+}
+\description{
+Update a manifest stage
+}
+\keyword{internal}
diff --git a/man/dot-manifest_start.Rd b/man/dot-manifest_start.Rd
new file mode 100644
index 0000000..078352e
--- /dev/null
+++ b/man/dot-manifest_start.Rd
@@ -0,0 +1,35 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.manifest_start}
+\alias{.manifest_start}
+\title{Start or load a dataset provenance manifest}
+\usage{
+.manifest_start(
+  manifest_path,
+  dataset_id,
+  duckdb_path,
+  base_dir = ".",
+  selection = list(),
+  hash_files = FALSE
+)
+}
+\arguments{
+\item{manifest_path}{Character. Path to the JSON manifest.}
+
+\item{dataset_id}{Character scalar.}
+
+\item{duckdb_path}{Character scalar.}
+
+\item{base_dir}{Character scalar.}
+
+\item{selection}{Optional named list describing the dataset selection.}
+
+\item{hash_files}{Logical. Calculate SHA-256 for manifest-recorded files.}
+}
+\value{
+A manifest object with \code{path} and \code{run_index}.
+}
+\description{
+Start or load a dataset provenance manifest
+}
+\keyword{internal}
diff --git a/man/dot-parseHMMEROutput.Rd b/man/dot-parseHMMEROutput.Rd
index 5ace7e1..4b95caf 100644
--- a/man/dot-parseHMMEROutput.Rd
+++ b/man/dot-parseHMMEROutput.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/helpers.R
 \name{.parseHMMEROutput}
 \alias{.parseHMMEROutput}
 \title{Parse HMMER tabular output into a tibble}
diff --git a/man/dot-parse_hmmer_profiles.Rd b/man/dot-parse_hmmer_profiles.Rd
index 428796e..12f056e 100644
--- a/man/dot-parse_hmmer_profiles.Rd
+++ b/man/dot-parse_hmmer_profiles.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/helpers.R
 \name{.parse_hmmer_profiles}
 \alias{.parse_hmmer_profiles}
 \title{Parsing HMM database to extract profile names, accessions and descriptions}
diff --git a/man/dot-pick_shell.Rd b/man/dot-pick_shell.Rd
index fb4e8bf..694f49a 100644
--- a/man/dot-pick_shell.Rd
+++ b/man/dot-pick_shell.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_curation.R
+% Please edit documentation in R/helpers.R
 \name{.pick_shell}
 \alias{.pick_shell}
 \title{Helps run a shell inside a container, and prefers bash (don't we all?)}
diff --git a/man/dot-prepareHmmerDatabases.Rd b/man/dot-prepareHmmerDatabases.Rd
index f0879b4..9f8f0d1 100644
--- a/man/dot-prepareHmmerDatabases.Rd
+++ b/man/dot-prepareHmmerDatabases.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/data_processing.R
 \name{.prepareHmmerDatabases}
 \alias{.prepareHmmerDatabases}
 \title{Download and prepare HMMER databases for generating new file types.}
diff --git a/man/dot-preserve_export_id_text.Rd b/man/dot-preserve_export_id_text.Rd
new file mode 100644
index 0000000..fea92d8
--- /dev/null
+++ b/man/dot-preserve_export_id_text.Rd
@@ -0,0 +1,16 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/helpers.R
+\name{.preserve_export_id_text}
+\alias{.preserve_export_id_text}
+\title{A helper used in data_curation.R and data_processing.R to ensure exported tables
+don't lose trailing zeroes. Should be relocated into a common helpers/utilities
+script later.}
+\usage{
+.preserve_export_id_text(df)
+}
+\description{
+A helper used in data_curation.R and data_processing.R to ensure exported tables
+don't lose trailing zeroes. Should be relocated into a common helpers/utilities
+script later.
+}
+\keyword{internal}
diff --git a/man/dot-processPanaroo.Rd b/man/dot-processPanaroo.Rd
index aa42f1f..f3aa5b7 100644
--- a/man/dot-processPanaroo.Rd
+++ b/man/dot-processPanaroo.Rd
@@ -12,7 +12,8 @@
   cluster_threshold,
   family_seq_identity,
   panaroo_threads_per_job,
-  refind_mode = c("off", "default", "strict")
+  refind_mode = c("off", "default", "strict"),
+  verbose = TRUE
 )
 }
 \arguments{
diff --git a/man/dot-process_chunk.Rd b/man/dot-process_chunk.Rd
deleted file mode 100644
index b6c9c21..0000000
--- a/man/dot-process_chunk.Rd
+++ /dev/null
@@ -1,45 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_processing.R
-\name{.process_chunk}
-\alias{.process_chunk}
-\title{Run InterProScan on a sequence chunk inside Docker}
-\usage{
-.process_chunk(
-  chunk,
-  path,
-  ipr_data_path = "inst/extdata/interpro/data",
-  out_file_base,
-  appl,
-  chunk_id,
-  threads,
-  file_format,
-  docker_image = sprintf("interpro/interproscan:\%s", "5.76-107.0")
-)
-}
-\arguments{
-\item{chunk}{A tibble with columns \code{name} and \code{sequence}.}
-
-\item{path}{Working directory used for temporary files.}
-
-\item{ipr_data_path}{Path to InterProScan data directory.}
-
-\item{out_file_base}{Output prefix for chunk results.}
-
-\item{appl}{Character vector of InterProScan applications (e.g. \code{"Pfam"}).}
-
-\item{chunk_id}{Integer chunk index.}
-
-\item{threads}{Number of CPUs for InterProScan container.}
-
-\item{file_format}{Output format (\code{"TSV"}).}
-
-\item{docker_image}{InterProScan Docker image.}
-}
-\value{
-Path to a \code{.tsv} or \code{.tsv.gz} InterProScan output file.
-}
-\description{
-Executes InterProScan on a subset of protein sequences, writing temporary
-FASTA and reading back \code{.tsv} or \code{.tsv.gz} results.
-}
-\keyword{internal}
diff --git a/man/dot-proteinAnnotations2Duckdb.Rd b/man/dot-proteinAnnotations2Duckdb.Rd
index e59930d..cfffa09 100644
--- a/man/dot-proteinAnnotations2Duckdb.Rd
+++ b/man/dot-proteinAnnotations2Duckdb.Rd
@@ -1,12 +1,13 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/data_processing.R
 \name{.proteinAnnotations2Duckdb}
 \alias{.proteinAnnotations2Duckdb}
 \title{Map HMMER protein annotations to genome-level count matrix and load into DuckDB}
 \usage{
 .proteinAnnotations2Duckdb(
   duckdb_path,
-  databases = c("Pfam", "COG", "AMRFinder")
+  databases,
+  output_path = dirname(duckdb_path)
 )
 }
 \arguments{
diff --git a/man/dot-readIPRscanTsv.Rd b/man/dot-readIPRscanTsv.Rd
deleted file mode 100644
index d4503e5..0000000
--- a/man/dot-readIPRscanTsv.Rd
+++ /dev/null
@@ -1,19 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_processing.R
-\name{.readIPRscanTsv}
-\alias{.readIPRscanTsv}
-\title{Internal helpers for reading InterProScan TSV outputs}
-\usage{
-.readIPRscanTsv(filepath)
-}
-\arguments{
-\item{filepath}{Path to a \code{.tsv} or \code{.tsv.gz} InterProScan result file.}
-}
-\value{
-A tibble of parsed InterProScan output.
-}
-\description{
-Provide standardized column names, types, and a reader wrapper for the
-InterProScan tab-delimited output format.
-}
-\keyword{internal}
diff --git a/man/dot-runHMMER.Rd b/man/dot-runHMMER.Rd
index 2ead2ac..72618a7 100644
--- a/man/dot-runHMMER.Rd
+++ b/man/dot-runHMMER.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/data_processing.R
 \name{.runHMMER}
 \alias{.runHMMER}
 \title{Wrapper for preparing HMM databases and running HMMER on protein sequences from duckdb and writing them.}
@@ -16,6 +16,23 @@
   verbose = TRUE
 )
 }
+\arguments{
+\item{duckdb_path}{path to the duckdb with protein sequences and list}
+
+\item{output_path}{path where HMMER output will be saved}
+
+\item{threads}{number of threads}
+
+\item{hmmer_db_dir}{path to the directory where HMM databases are/will be downloaded}
+
+\item{databases}{list of HMM databases}
+
+\item{docker_image}{the docker image of HMMER}
+
+\item{num_of_splits}{The number of splits of the protein sequence file for parallel processing}
+
+\item{n_workers}{The number of parallel runs}
+}
 \description{
 Wrapper for preparing HMM databases and running HMMER on protein sequences from duckdb and writing them.
 }
diff --git a/man/dot-runHmmerJob.Rd b/man/dot-runHmmerJob.Rd
index 5cc8340..121d513 100644
--- a/man/dot-runHmmerJob.Rd
+++ b/man/dot-runHmmerJob.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/data_processing.R
 \name{.runHmmerJob}
 \alias{.runHmmerJob}
 \title{The function to run HMMER with docker}
@@ -17,6 +17,28 @@
   verbose = TRUE
 )
 }
+\arguments{
+\item{JOB_NAME}{protein_chunk id}
+
+\item{FASTA}{fasta sequences}
+
+\item{DB}{HMM database}
+
+\item{Total_proteins}{protein sequence count}
+
+\item{output_path}{path for saving hmmer outputs}
+
+\item{db_paths}{path to HMM database}
+
+\item{docker_image}{hmmer docker image (ideally from dockerhub)}
+
+\item{threads}{number of threads}
+
+\item{n_workers}{number of parallel workers}
+}
+\value{
+the filename of the parquet file with hmmer output post parsing
+}
 \description{
 The function to run HMMER with docker
 }
diff --git a/man/dot-runPanaroo.Rd b/man/dot-runPanaroo.Rd
index ed083fa..c302921 100644
--- a/man/dot-runPanaroo.Rd
+++ b/man/dot-runPanaroo.Rd
@@ -16,7 +16,8 @@
   refind_mode = c("off", "default", "strict"),
   strip_pseudogenes = FALSE,
   pseudogene_clean_dir = "gff_clean",
-  write_pseudogene_audit = TRUE
+  write_pseudogene_audit = TRUE,
+  verbose = TRUE
 )
 }
 \arguments{
diff --git a/man/dot-stripPseudogeneGFFs.Rd b/man/dot-stripPseudogeneGFFs.Rd
index f2e4471..3853720 100644
--- a/man/dot-stripPseudogeneGFFs.Rd
+++ b/man/dot-stripPseudogeneGFFs.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_processing.R
+% Please edit documentation in R/helpers.R
 \name{.stripPseudogeneGFFs}
 \alias{.stripPseudogeneGFFs}
 \title{Remove pseudogene annotations from Panaroo input GFF files}
diff --git a/man/dot-to_container.Rd b/man/dot-to_container.Rd
index ce8b31c..c12cc21 100644
--- a/man/dot-to_container.Rd
+++ b/man/dot-to_container.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/data_processing.R
+% Please edit documentation in R/helpers.R
 \name{.to_container}
 \alias{.to_container}
 \title{.to_container()}
diff --git a/man/dot-write_compressed_parquet.Rd b/man/dot-write_compressed_parquet.Rd
index b9e45de..586cf50 100644
--- a/man/dot-write_compressed_parquet.Rd
+++ b/man/dot-write_compressed_parquet.Rd
@@ -1,5 +1,5 @@
 % Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/runHMMER.R
+% Please edit documentation in R/helpers.R
 \name{.write_compressed_parquet}
 \alias{.write_compressed_parquet}
 \title{Write a data frame to a compressed Parquet file}
diff --git a/man/exportProcessedData.Rd b/man/exportProcessedData.Rd
index c8f16d4..07aff2a 100644
--- a/man/exportProcessedData.Rd
+++ b/man/exportProcessedData.Rd
@@ -11,6 +11,7 @@ exportProcessedData(
   export_formats = c("csv"),
   export_sequences = FALSE,
   tables = NULL,
+  export_tables = TRUE,
   verbose = TRUE
 )
 }
diff --git a/man/exportTables.Rd b/man/exportTables.Rd
index 4fffb78..98618a7 100644
--- a/man/exportTables.Rd
+++ b/man/exportTables.Rd
@@ -8,8 +8,9 @@ exportTables(
   duckdb_path,
   output_dir = NULL,
   tables = NULL,
-  skip_tables = c(NULL),
+  skip_tables = NULL,
   include_summary = TRUE,
+  export_formats = c("csv"),
   export_tables = TRUE,
   load_tables = FALSE,
   verbose = TRUE
diff --git a/man/runDataProcessing.Rd b/man/runDataProcessing.Rd
index fa2eda4..4a80dc7 100644
--- a/man/runDataProcessing.Rd
+++ b/man/runDataProcessing.Rd
@@ -2,7 +2,7 @@
 % Please edit documentation in R/data_processing.R
 \name{runDataProcessing}
 \alias{runDataProcessing}
-\title{Run the full amRdata processing pipeline (Panaroo → CD-HIT → InterProScan → Parquet)}
+\title{Run the full amRdata processing pipeline (Panaroo -> CD-HIT -> HMMER -> Parquet)}
 \usage{
 runDataProcessing(
   duckdb_path,
@@ -22,12 +22,11 @@ runDataProcessing(
   cdhit_memory = 0,
   cdhit_extra_args = c("-g", "1"),
   cdhit_output_prefix = "cdhit_out",
-  ipr_appl = c("Pfam"),
-  ipr_threads_unused = NULL,
-  ipr_version = "5.76-107.0",
-  ipr_dest_dir = "inst/extdata/interpro",
-  ipr_platform = "linux/amd64",
-  auto_prepare_data = TRUE,
+  hmmer_databases = c("Pfam", "COG", "AMRFinder", "DefenseCas"),
+  hmmer_db_dir = NULL,
+  hmmer_docker_image = "staphb/hmmer",
+  hmmer_num_splits = 8L,
+  hmmer_workers = 8L,
   ref_file_path = "data_raw/",
   verbose = TRUE
 )
@@ -35,17 +34,19 @@ runDataProcessing(
 \arguments{
 \item{duckdb_path}{Character. Path to the \strong{per-selection DuckDB} produced by
 \code{\link[=prepareGenomes]{prepareGenomes()}} (e.g., \code{"data/<Bug>/<Abbrev>.duckdb"}). This DB must
-already contain at least the tables written by \code{prepareGenomes()} and subsequent
-download steps (e.g., \code{files}, \code{filtered}, and metadata tables).}
+already contain the tables written by \code{\link[=prepareGenomes]{prepareGenomes()}} and the upstream
+genome-processing steps.}
 
-\item{output_path}{Character or \code{NULL}. Base directory for writing Panaroo/CD-HIT/InterProScan
-outputs and final Parquet files. If \code{NULL}, defaults to \code{dirname(duckdb_path)}.}
+\item{output_path}{Character or \code{NULL}. Base directory for writing Panaroo,
+CD-HIT, HMMER, and final Parquet outputs. If \code{NULL}, defaults to
+\code{dirname(duckdb_path)}.}
 
-\item{threads}{Integer. Shared concurrency budget used across tools (Panaroo, CD-HIT,
-InterProScan). Passed through to each stage as appropriate. Defaults to \code{8}.}
+\item{threads}{Integer. Shared concurrency budget used across Panaroo, CD-HIT,
+and HMMER. Defaults to \code{8}.}
 
-\item{panaroo_split_jobs}{Logical. If \code{TRUE}, Panaroo runs in multiple batches that can be
-merged by \code{\link[=.mergePanaroo]{.mergePanaroo()}}. If \code{FALSE}, Panaroo runs once on all isolates. Default: \code{FALSE}.}
+\item{panaroo_split_jobs}{Logical. If \code{TRUE}, Panaroo runs in multiple batches
+that can be merged by \code{\link[=.mergePanaroo]{.mergePanaroo()}}. If \code{FALSE}, Panaroo runs once on all
+isolates. Default: \code{FALSE}.}
 
 \item{panaroo_core_threshold}{Numeric. Panaroo \code{--core_threshold}. Default: \code{0.90}.}
 
@@ -53,139 +54,145 @@ merged by \code{\link[=.mergePanaroo]{.mergePanaroo()}}. If \code{FALSE}, Panaro
 
 \item{panaroo_cluster_threshold}{Numeric. Panaroo \code{--threshold}. Default: \code{0.95}.}
 
-\item{panaroo_family_seq_identity}{Numeric. Panaroo \code{-f} (gene family identity). Default: \code{0.5}.}
+\item{panaroo_family_seq_identity}{Numeric. Panaroo \code{-f} gene family identity.
+Default: \code{0.5}.}
 
-\item{panaroo_refind_mode}{Character. Panaroo's \code{--refind-mode} (\code{"off"}, \code{"default"},
-or \code{"strict"}). See \code{\link[=.processPanaroo]{.processPanaroo()}} for what refinding does and the runtime
-caveat behind the current default. Default \code{"off"}.}
+\item{panaroo_refind_mode}{Character. Panaroo's \code{--refind-mode} (\code{"off"},
+\code{"default"}, or \code{"strict"}). See \code{\link[=.processPanaroo]{.processPanaroo()}} for the runtime caveat
+behind refinding. Default: \code{"off"}.}
+
+\item{panaroo_strip_pseudogenes}{Logical. If \code{TRUE}, remove pseudogene feature
+records from Panaroo input GFF files before running Panaroo. Default: \code{FALSE}.}
+
+\item{panaroo_pseudogene_clean_dir}{Character. Directory name for cleaned GFF
+files. Default: \code{"gff_clean"}.}
+
+\item{panaroo_write_pseudogene_audit}{Logical. If \code{TRUE}, write a pseudogene
+cleaning audit file. Default: \code{TRUE}.}
 
 \item{cdhit_identity}{Numeric. CD-HIT \code{-c} identity threshold. Default: \code{0.9}.}
 
 \item{cdhit_word_length}{Integer. CD-HIT \code{-n} word length. Default: \code{5}.}
 
-\item{cdhit_memory}{Integer. CD-HIT \code{-M} memory limit (MB). Use \code{0} for unlimited. Default: \code{0}.}
+\item{cdhit_memory}{Integer. CD-HIT \code{-M} memory limit in MB. Use \code{0} for
+unlimited. Default: \code{0}.}
 
-\item{cdhit_extra_args}{Character vector. Extra arguments forwarded to \code{cd-hit}
-(e.g., \code{c("-g","1")}). Default: \code{c("-g","1")}.}
+\item{cdhit_extra_args}{Character vector. Extra arguments forwarded to
+\code{cd-hit}. Default: \code{c("-g", "1")}.}
 
-\item{cdhit_output_prefix}{Character. Prefix for CD-HIT output files. Default: \code{"cdhit_out"}.}
+\item{cdhit_output_prefix}{Character. Prefix for CD-HIT output files.
+Default: \code{"cdhit_out"}.}
 
-\item{ipr_appl}{Character vector. InterProScan applications to run; typically \code{c("Pfam")}.
-Default: \code{c("Pfam")}.}
+\item{hmmer_databases}{Character vector. HMMER annotation databases to run.
+Default: \code{c("Pfam", "COG", "AMRFinder", "DefenseCas")}.}
 
-\item{ipr_threads_unused}{Deprecated/unused. Kept for backward compatibility; ignored.}
+\item{hmmer_db_dir}{Character or \code{NULL}. Directory containing the shared HMMER
+database cache. If \code{NULL}, uses the amRdata user cache.}
 
-\item{ipr_version}{Character. InterProScan image tag (e.g., \code{"5.76-107.0"}). Default: \code{"5.76-107.0"}.}
+\item{hmmer_docker_image}{Character. Docker image containing HMMER.
+Default: \code{"staphb/hmmer"}.}
 
-\item{ipr_dest_dir}{Character. Local destination for InterProScan data bundle
-(used by \code{.checkInterProData()}). Default: \code{"inst/extdata/interpro"}.}
+\item{hmmer_num_splits}{Integer. Number of protein-sequence chunks for HMMER.
+Default: \code{8}.}
 
-\item{ipr_platform}{Character. Docker platform string for InterProScan containers,
-e.g., \code{"linux/amd64"}. Default: \code{"linux/amd64"}.}
+\item{hmmer_workers}{Integer. Number of parallel HMMER workers. Default: \code{8}.}
 
-\item{auto_prepare_data}{Logical. If \code{TRUE}, ensure InterProScan data are present
-(download/verify if missing). Default: \code{TRUE}.}
-
-\item{ref_file_path}{Character. Directory containing reference TSVs used by cleanData()
-for metadata harmonization (e.g., \code{"data_raw/"}). \strong{Required}; defaults to \code{"data_raw/"}.}
+\item{ref_file_path}{Character. Directory containing reference TSVs used by
+\code{\link[=cleanMetaData]{cleanMetaData()}} and \code{\link[=cleanData]{cleanData()}} for metadata harmonization.
+Default: \code{"data_raw/"}.}
 
 \item{verbose}{Logical. Print progress messages. Default: \code{TRUE}.}
 }
 \value{
 Invisibly returns a list with:
 \itemize{
-\item \code{duckdb_path} – input DuckDB path
-\item \code{panaroo_output} – path to the selected Panaroo output directory used for import
-\item \code{parquet_duckdb_path} – absolute path to the created Parquet-backed DuckDB
+\item \code{duckdb_path} - input DuckDB path
+\item \code{panaroo_output} - path to the selected Panaroo output directory used for import
+\item \code{parquet_duckdb_path} - absolute path to the created Parquet-backed DuckDB
 }
 }
 \description{
 \code{runDataProcessing()} orchestrates the complete feature-extraction pipeline for a
-BV-BRC selection, starting from a \strong{per-selection DuckDB} (created by
-\code{\link[=prepareGenomes]{prepareGenomes()}} and populated by downstream steps). It:
+BV-BRC selection, starting from a \strong{per-selection DuckDB} created by
+\code{\link[=prepareGenomes]{prepareGenomes()}} and populated by downstream genome processing steps. It:
 \enumerate{
 \item Runs \strong{Panaroo} to build the pangenome and writes gene/struct outputs into DuckDB.
 \item Runs \strong{CD-HIT} to cluster proteins and writes protein outputs into DuckDB.
-\item Runs \strong{InterProScan} (Pfam) to annotate protein domains and writes domain outputs into DuckDB.
+\item Runs \strong{HMMER} against the requested protein databases and writes annotation
+tables into DuckDB.
 \item \strong{Cleans BV-BRC metadata} (drug names/classes, countries, years) and
-exports all feature/metadata tables as \strong{compressed Parquet} files, then creates
-a \strong{Parquet-backed DuckDB} with read-only views of those Parquets for downstream ML.
+exports feature and metadata tables as compressed Parquet files, then creates
+a \strong{Parquet-backed DuckDB} with read-only views for downstream ML.
 }
 
 The function is a thin controller that delegates each stage to the corresponding
-internal helpers (Dockerized tools where applicable) and ensures consistent
-output locations and table schemas across stages.
+internal helpers (Dockerized tools where applicable) and records processing
+parameters and provenance in the dataset manifest.
 }
 \details{
 \strong{Docker & Platform Notes}
 \itemize{
-\item All heavy tools (Panaroo, CD-HIT, InterProScan) run inside Docker containers.
-\item On Apple Silicon/ARM hosts, images are forced to \verb{--platform linux/amd64} to ensure compatibility.
-\item Ensure Docker Desktop is running and has sufficient memory/CPUs configured.
+\item Panaroo, CD-HIT, and HMMER run inside Docker containers.
+\item HMMER databases are stored separately from individual bug directories and
+are reused across datasets unless a custom \code{hmmer_db_dir} is supplied.
+\item Ensure Docker Desktop is running and has sufficient memory and CPU resources.
 }
 
 \strong{Input Requirements}
 \itemize{
-\item The \code{duckdb_path} must reference a per-selection DuckDB that contains:
-\code{files} (paths to \code{.gff}, \code{.fna}, \code{.PATRIC.faa}),
-\code{filtered} (genomes selected for download/filtering), and
-BV-BRC metadata tables written by earlier steps.
+\item \code{duckdb_path} must reference a per-selection DuckDB containing the genome
+file table, filtered genome selection, and BV-BRC metadata produced by the
+upstream curation workflow.
 }
 
 \strong{Outputs & Side Effects}
 \itemize{
-\item Writes tool-specific intermediate outputs under \code{output_path} (e.g., \verb{panaroo_out_*}, CD-HIT files).
-\item Writes Parquet files to \code{output_path}:
-\code{gene_count.parquet}, \code{protein_count.parquet}, \code{domain_count.parquet}, \code{struct.parquet},
-\code{gene_names.parquet}, \code{protein_names.parquet}, \code{domain_names.parquet},
-\code{gene_seqs.parquet}, \code{protein_seqs.parquet}, \code{genome_gene_protein.parquet},
-\code{metadata.parquet}, \code{amr_phenotype.parquet}, \code{genome_data.parquet}, \code{original_metadata.parquet}.
-\item Creates a new Parquet-backed DuckDB (\verb{*_parquet.duckdb}) with read-only views pointing to those Parquets.
+\item Writes tool-specific intermediate outputs under \code{output_path}.
+\item Writes feature and metadata Parquet files under \code{output_path}.
+\item Creates a new Parquet-backed DuckDB (\verb{*_parquet.duckdb}) with read-only views
+over the generated Parquet files.
+\item Records processing parameters, software versions, database selections, and
+other provenance information in the dataset manifest.
 }
 
 \strong{Threading}
 \itemize{
-\item \code{threads} is a shared budget; each stage uses a portion or all of it.
-\item InterProScan can be memory-intensive; on laptops, single-container mode is used internally.
+\item \code{threads} provides the shared CPU budget for the major processing stages.
+\item Panaroo, CD-HIT, and HMMER allocate that budget according to their respective
+stage parameters.
 }
 }
 \section{Pipeline Steps}{
 
 \enumerate{
-\item \strong{Panaroo} via runPanaroo2Duckdb() → writes:
+\item \strong{Panaroo} via \code{\link[=runPanaroo2Duckdb]{runPanaroo2Duckdb()}} -> writes:
 \itemize{
-\item \code{gene_count} (genome × gene counts)\cr
+\item \code{gene_count} (genome x gene counts)\cr
 \item \code{gene_names}\cr
 \item \code{gene_struct} (structural variants)\cr
 \item \code{gene_ref_seq}, \code{genome_gene_protein}
 }
-\item \strong{CD-HIT} via CDHIT2duckdb() (calls internal \code{.runCDHIT()}) → writes:
+\item \strong{CD-HIT} via \code{\link[=CDHIT2duckdb]{CDHIT2duckdb()}} -> writes:
 \itemize{
-\item \code{protein_count} (genome × protein-cluster counts)\cr
+\item \code{protein_count} (genome x protein-cluster counts)\cr
 \item \code{protein_names}\cr
-\item \code{protein_cluster_seq} (representative sequences)
-}
-\item \strong{InterProScan (Pfam)} via domainFromIPR() → writes:
-\itemize{
-\item \code{domain_names}\cr
-\item \code{domain_count} (genome × domain-family matrix)
+\item \code{protein_cluster_seq} (representative sequences)\cr
+\item \code{protein_members}
 }
-\item \strong{Metadata cleaning + Parquet export} via cleanData() → writes Parquet
-files to \code{output_path}, and builds a \strong{Parquet-backed DuckDB}
-(\verb{*_parquet.duckdb}) with views:
+\item \strong{HMMER} via the configured HMMER databases -> writes:
 \itemize{
-\item \code{gene_count}, \code{protein_count}, \code{domain_count}, \code{struct}\cr
-\item \code{metadata} (cleaned), plus \code{amr_phenotype}, \code{genome_data}, \code{original_metadata}\cr
-\item \code{gene_names}, \code{protein_names}, \code{domain_names}\cr
-\item \code{gene_seqs}, \code{protein_seqs}\cr
-\item \code{genome_gene_protein}
+\item \verb{protein_<database>} annotation tables\cr
+\item \verb{protein_<database>_count} genome-by-annotation count tables
+\item The default databases are \code{Pfam}, \code{COG}, \code{AMRFinder}, and \code{DefenseCas}.
 }
+\item \strong{Metadata cleaning + Parquet export} via \code{\link[=cleanData]{cleanData()}} -> writes
+Parquet files to \code{output_path}, and builds a \strong{Parquet-backed DuckDB}
+(\verb{*_parquet.duckdb}) with views over those Parquets.
 }
 }
 
 \examples{
 \dontrun{
-# Paths below are illustrative; adapt to your project layout.
 runDataProcessing(
   duckdb_path   = "data/Shigella_flexneri/Sfl.duckdb",
   output_path   = "data/Shigella_flexneri",
@@ -194,11 +201,12 @@ runDataProcessing(
 )
 
 # After completion:
-#   data/Shigella_flexneri/Sfl_parquet.duckdb
+# data/Shigella_flexneri/Sfl_parquet.duckdb
 # will contain views over the Parquet files for downstream ML.
 }
 
 }
 \seealso{
-prepareGenomes(), runPanaroo2Duckdb(), CDHIT2duckdb(), domainFromIPR(), cleanData()
+\code{\link[=prepareGenomes]{prepareGenomes()}}, \code{\link[=runPanaroo2Duckdb]{runPanaroo2Duckdb()}}, \code{\link[=CDHIT2duckdb]{CDHIT2duckdb()}}, \code{\link[=cleanMetaData]{cleanMetaData()}},
+\code{\link[=cleanData]{cleanData()}}
 }
diff --git a/man/runPanaroo2Duckdb.Rd b/man/runPanaroo2Duckdb.Rd
index 48bec57..38f2b51 100644
--- a/man/runPanaroo2Duckdb.Rd
+++ b/man/runPanaroo2Duckdb.Rd
@@ -131,6 +131,6 @@ runPanaroo2Duckdb(
 \item \code{.runPanaroo()} — core Panaroo execution
 \item \code{.mergePanaroo()} — merge multiple Panaroo batches
 \item \code{.panaroo2duckdb()} — import Panaroo results into DuckDB
-\item \code{\link[=runDataProcessing]{runDataProcessing()}} — full pipeline including CD-HIT & InterProScan
+\item \code{\link[=runDataProcessing]{runDataProcessing()}} — full pipeline including CD-HIT & HMMER
 }
 }
diff --git a/tests/testthat/test-buildDyadFeatureMap.R b/tests/testthat/test-buildDyadFeatureMap.R
new file mode 100644
index 0000000..04dedd9
--- /dev/null
+++ b/tests/testthat/test-buildDyadFeatureMap.R
@@ -0,0 +1,23 @@
+test_that("buildDyadFeatureMap() errors when the DuckDB path does not exist", {
+  expect_error(
+    buildDyadFeatureMap(duckdb_path = tempfile(fileext = ".duckdb"))
+  )
+})
+
+test_that("buildDyadFeatureMap() errors when no provenance manifest is found", {
+  tmp_dir <- file.path(tempdir(), paste0("dyad-test-", as.integer(runif(1, 1, 1e6))))
+  dir.create(tmp_dir)
+  on.exit(unlink(tmp_dir, recursive = TRUE), add = TRUE)
+
+  fake_db <- file.path(tmp_dir, "fake.duckdb")
+  file.create(fake_db)
+
+  expect_error(
+    buildDyadFeatureMap(duckdb_path = fake_db),
+    regexp = "provenance manifest"
+  )
+})
+
+# TODO (Bioconductor): add an end-to-end test that runs buildDyadFeatureMap()
+# against a small fixture DuckDB + manifest + annotation Parquets and checks the
+# resulting edge list (source/target columns, dyad format, type prefixes).