v3.5.0 release candidate

.gitignore

@@ -15,8 +15,8 @@
 !pixi.lock
 !pyproject.toml
 !uv.lock
-!.envrc
-!flake.nix
+!mise.lock
+!mise.toml
 
 # workflows
 !/workflows

Containerfile

@@ -40,7 +40,7 @@ ENV PATH="${HOME}/.pixi/bin:${PATH}"
 # environment, so skip those direct packages and their dependency subtrees.
 RUN cd $HOME && \
     pixi install --frozen \
-        --skip-with-deps apptainer \
+        --skip apptainer \
         --skip-with-deps rust-script \
         --skip-with-deps rust \
         --skip-with-deps compilers \

README.md

@@ -212,6 +212,8 @@ Then generate a samplesheet from that accession list:
 nvd samplesheet generate --from-sra accessions.txt --platform illumina --output samplesheet.csv
 ```
 
+NVD decodes each SRA run as a stream through target enrichment rather than materializing decoded raw FASTQ files. Raw-read FastQC therefore runs only for local FASTQ inputs; the `skip_fastqc` setting controls those local tasks. SRA reads rejoin local reads immediately after target enrichment and receive the same subsequent preprocessing.
+
 If you want to inspect what NVD would write before touching the filesystem, use dry-run mode:
 
 ```bash

bin/build_sequence_flow.py

@@ -181,22 +181,6 @@ def megablast_partition_rows(path: Path) -> list[LedgerRow]:
     return rows
 
 
-def blast_filter_rows(path: Path) -> list[LedgerRow]:
-    return [
-        LedgerRow(
-            sample_id=source["sample_id"],
-            stage=source["stage"],
-            input_class=source["query_class"],
-            output_class="virus_only_hits",
-            decision="retain",
-            reason="viral_taxonomy_match",
-            sequences_in=source["queries_in"],
-            sequences_out=source["queries_retained"],
-        )
-        for source in read_tsv(path)
-    ]
-
-
 def rows_for_input(path: Path) -> list[LedgerRow]:
     name = path.name
     adapters = (
@@ -206,10 +190,6 @@ def rows_for_input(path: Path) -> list[LedgerRow]:
         (("_mapped_counts.txt",), mapback_rows),
         ((".blast_query_batches.tsv",), query_batch_rows),
         ((".megablast_query_partition.tsv",), megablast_partition_rows),
-        (
-            (".megablast_query_filtering.tsv", ".blastn_query_filtering.tsv"),
-            blast_filter_rows,
-        ),
     )
     for suffixes, adapter in adapters:
         if name.endswith(suffixes):

bin/finalize_blast_results.py

@@ -98,7 +98,7 @@ def main(argv: list[str] | None = None) -> None:
     parser.add_argument(
         "--virus-index-version",
         required=True,
-        help="STAT k-mer database version used for virus enrichment index",
+        help="Target-enrichment index version",
     )
     parser.add_argument(
         "--run-id",

bin/stream_fastqs_to_deacon.py

@@ -42,6 +42,7 @@ class DeaconStreamConfig:
     rel_threshold: float
     deplete: bool
     deacon_bin: str
+    check_pairs: bool = False
 
 
 @dataclass(frozen=True)
@@ -66,6 +67,7 @@ def parse_args(argv: list[str] | None = None) -> argparse.Namespace:
     parser.add_argument("--abs-threshold", type=int, default=1)
     parser.add_argument("--rel-threshold", type=float, default=0.0)
     parser.add_argument("--deplete", action="store_true")
+    parser.add_argument("--check-pairs", action="store_true")
     parser.add_argument("--deacon-bin", default="deacon")
     return parser.parse_args(argv)
 
@@ -94,6 +96,7 @@ def config_from_args(args: argparse.Namespace) -> DeaconStreamConfig:
         rel_threshold=args.rel_threshold,
         deplete=args.deplete,
         deacon_bin=args.deacon_bin,
+        check_pairs=args.check_pairs,
     )
 
 
@@ -144,6 +147,8 @@ def validate_config(
     if single == paired:
         message = "Provide either --reads-list for single-end input or both --r1-list and --r2-list for paired input"
         raise StreamError(message)
+    if config.check_pairs and not paired:
+        raise StreamError("check-pairs requires paired input")
     if single:
         validate_files("reads", config.reads)
         validate_sample_compression(config.reads)
@@ -204,9 +209,11 @@ def stream_files_to_fifo(
 
 
 def deacon_command(config: DeaconStreamConfig, inputs: tuple[Path, ...]) -> list[str]:
+    pair_args = ["--check-pairs"] if config.check_pairs and config.r1 else []
     command = [
         config.deacon_bin,
         "filter",
+        *pair_args,
         "--threads",
         str(config.threads),
         "--abs-threshold",