0.4.1

docs/source/tutorials/cli_usage.md

@@ -29,7 +29,7 @@ The load command builds the raw AnnData object from your raw sequencing data. It
 - adata.X contains binarized modification data (conversion/deaminase), or modification probabilitiesc (native).
 - Adds basic read-level QC annotations (Read start, end, length, mean quality).
 - Adds layers encoding read DNA sequences, base quality scores, base mismatches.
-- Maintains BAM Tags/Flags in adata.obs.
+- Maintains BAM tags/flags in adata.obs (UMI and barcode annotations loaded from Parquet sidecars).
 - Writes the raw AnnData to the canonical output path and runs MultiQC.
 - Optionally deletes intermediate BAMs, H5ADs, and TSVs.
 

docs/source/tutorials/experiment_config.md

@@ -57,18 +57,25 @@ Below are some of the most commonly edited fields and how they affect the CLI wo
 - Lists are written in bracketed form, e.g. `[5mC]` or `[5mC_5hmC]`.
 - If you update the CSV, re-run the CLI command pointing at the updated file.
 
-## BAM tags
+## Read annotations
 
-smftools writes and/or propagates the following BAM tags when loading data. These are also loaded
-into `adata.obs` when `load_adata` reads BAM tags.
+smftools annotates reads during `load_adata` and stores the results in `adata.obs`. Standard BAM
+tags (e.g. `NM`, `MD`, `MM`, `ML`) are read directly from BAM files. UMI and barcode annotations
+are computed in parallel and written to Parquet sidecar files alongside the aligned BAM, then loaded
+into `adata.obs` from those sidecars. The aligned BAM itself is not modified.
 
-**UMI tags**
+**UMI annotations** (written to `.umi_tags.parquet`)
 
-- `U1`: UMI from the *top* flank (read start or read end depending on match).
-- `U2`: UMI from the *bottom* flank.
+- `U1`: Orientation-corrected UMI for the *left* reference end of the mapped fragment (forward reads: US, reverse reads: UE).
+- `U2`: Orientation-corrected UMI for the *right* reference end of the mapped fragment (forward reads: UE, reverse reads: US).
+- `US`: Positional UMI from read start (delimited `UMI_seq;slot;flank_seq`).
+- `UE`: Positional UMI from read end (delimited `UMI_seq;slot;flank_seq`).
 - `RX`: Combined UMI string (`U1-U2`, or `U1`/`U2` if only one is present).
+- `FC`: Flank context of the U1/U2 pair (e.g. `top-bottom`).
 
-**Barcode tags (smftools demux backend)**
+When `threads` is set, UMI extraction is parallelized across multiple CPU cores.
+
+**Barcode annotations (smftools demux backend)** (written to `.barcode_tags.parquet`)
 
 - `BC`: Assigned barcode name, or `unclassified`.
 - `BM`: Match type (`both`, `read_start_only`, `read_end_only`, `mismatch`, `unclassified`).
@@ -79,9 +86,13 @@ into `adata.obs` when `load_adata` reads BAM tags.
 - `B5`: Barcode name matched at the read start (corresponds to `B1`/`B3`).
 - `B6`: Barcode name matched at the read end (corresponds to `B2`/`B4`).
 
-**Barcode tags (dorado demux backend)**
+When `threads` is set, barcode extraction is parallelized across multiple CPU cores.
+Demultiplexing (splitting reads into per-barcode BAMs) uses the sidecar `BC` assignments.
+Only primary alignments are included in split BAMs and sidecar files.
+
+**Barcode annotations (dorado demux backend)**
 
-- `BC`: Assigned barcode name.
+- `BC`: Assigned barcode name (read from BAM tag).
 - `bi`: Dorado barcode info array (if present; expanded into columns during load).
 
 Notes:

src/smftools/cli/load_adata.py

@@ -6,6 +6,7 @@
 from typing import Iterable, Union
 
 import numpy as np
+import pandas as pd
 
 from smftools.constants import BARCODE_KIT_ALIASES, LOAD_DIR, LOGGING_DIR, UMI_KIT_ALIASES
 from smftools.logging_utils import get_logger, setup_logging
@@ -516,6 +517,8 @@ def load_adata_core(cfg, paths: AdataPaths, config_path: str | None = None):
     ########################################################################################################################
 
     ################################### 4.5) Optional UMI annotation #############################################
+    umi_sidecar = None
+    barcode_sidecar = None
     if getattr(cfg, "use_umi", False):
         logger.info("Annotating UMIs in aligned and sorted BAM before demultiplexing")
 
@@ -538,20 +541,20 @@ def load_adata_core(cfg, paths: AdataPaths, config_path: str | None = None):
             umi_kit_config = load_umi_config_from_yaml(umi_yaml_path)
         resolved_umi = resolve_umi_config(umi_kit_config, cfg)
 
-        annotate_umi_tags_in_bam(
+        umi_sidecar = annotate_umi_tags_in_bam(
             aligned_sorted_output,
             use_umi=True,
-            umi_adapters=getattr(cfg, "umi_adapters", None),
+            umi_kit_config=umi_kit_config,
             umi_length=getattr(cfg, "umi_length", None),
             umi_search_window=getattr(cfg, "umi_search_window", 200),
-            umi_adapter_matcher=getattr(cfg, "umi_adapter_matcher", "exact"),
+            umi_adapter_matcher=getattr(cfg, "umi_adapter_matcher", "edlib"),
             umi_adapter_max_edits=resolved_umi["umi_adapter_max_edits"],
             samtools_backend=cfg.samtools_backend,
-            umi_kit_config=umi_kit_config,
             umi_ends=resolved_umi["umi_ends"],
             umi_flank_mode=resolved_umi["umi_flank_mode"],
             umi_amplicon_max_edits=resolved_umi["umi_amplicon_max_edits"],
             same_orientation=resolved_umi.get("same_orientation", False),
+            threads=cfg.threads,
         )
     ########################################################################################################################
 
@@ -603,7 +606,7 @@ def load_adata_core(cfg, paths: AdataPaths, config_path: str | None = None):
         resolved_bc = resolve_barcode_config(barcode_kit_config, cfg)
 
         logger.info("Extracting and assigning barcodes to aligned BAM using smftools backend")
-        barcoded_bam = extract_and_assign_barcodes_in_bam(
+        barcode_sidecar = extract_and_assign_barcodes_in_bam(
             aligned_sorted_output,
             barcode_adapters=getattr(cfg, "barcode_adapters", [None, None]),
             barcode_references=barcode_references,
@@ -619,10 +622,9 @@ def load_adata_core(cfg, paths: AdataPaths, config_path: str | None = None):
             barcode_kit_config=barcode_kit_config,
             barcode_ends=resolved_bc["barcode_ends"],
             barcode_amplicon_gap_tolerance=resolved_bc["barcode_amplicon_gap_tolerance"],
+            threads=cfg.threads,
         )
-        # Update aligned_sorted_output to point to the barcoded BAM
-        aligned_sorted_output = barcoded_bam
-        logger.info(f"smftools barcode extraction complete: {barcoded_bam}")
+        logger.info(f"smftools barcode extraction complete: {barcode_sidecar}")
     ########################################################################################################################
 
     ################################### 5) Demultiplexing ######################################################################
@@ -646,6 +648,7 @@ def load_adata_core(cfg, paths: AdataPaths, config_path: str | None = None):
                 cfg.split_path,
                 cfg.bam_suffix,
                 samtools_backend=cfg.samtools_backend,
+                barcode_sidecar=barcode_sidecar,
             )
 
             unclassified_bams = [p for p in all_bam_files if "unclassified" in p.name]
@@ -716,7 +719,7 @@ def load_adata_core(cfg, paths: AdataPaths, config_path: str | None = None):
             # Annotate BM tag from bi per-end scores on each demuxed BAM
             for bam in bam_files:
                 if "unclassified" not in bam.name:
-                    annotate_demux_type_from_bi_tag(bam, threshold=0.0)
+                    annotate_demux_type_from_bi_tag(bam)
 
             se_bam_files = bam_files
             bam_dir = cfg.split_path
@@ -956,14 +959,10 @@ def load_adata_core(cfg, paths: AdataPaths, config_path: str | None = None):
     default_tags = ["NM", "MD", "fn"]
     if cfg.smf_modality == "direct":
         default_tags.extend(["MM", "ML"])
-    # Add UMI tags if UMI extraction was enabled
-    if getattr(cfg, "use_umi", False):
-        default_tags.extend(["U1", "U2", "RX"])
-    # Add barcode tags if smftools barcode extraction was used
-    if demux_backend == "smftools" and cfg.barcode_kit:
-        default_tags.extend(["BC", "BM", "B1", "B2", "B3", "B4", "B5", "B6"])
+    # UMI tags are loaded from Parquet sidecar below (not from BAM)
+    # Barcode tags are loaded from Parquet sidecar below (not from BAM)
     # Add barcode tags from dorado single-pass demux (BM annotated from bi tag)
-    elif demux_backend == "dorado" and cfg.barcode_kit and not cfg.input_already_demuxed:
+    if demux_backend == "dorado" and cfg.barcode_kit and not cfg.input_already_demuxed:
         dorado_ver = _get_dorado_version()
         if dorado_ver is not None and dorado_ver >= (1, 3, 1):
             default_tags.extend(["BC", "BM", "bi"])
@@ -980,6 +979,31 @@ def load_adata_core(cfg, paths: AdataPaths, config_path: str | None = None):
         samtools_backend=cfg.samtools_backend,
     )
 
+    # Load UMI tags from Parquet sidecar (written by annotate_umi_tags_in_bam)
+    if getattr(cfg, "use_umi", False) and umi_sidecar and Path(umi_sidecar).exists():
+        logger.info("Loading UMI tags from Parquet sidecar: %s", umi_sidecar)
+        umi_df = pd.read_parquet(umi_sidecar).set_index("read_name")
+        umi_df = umi_df.reindex(raw_adata.obs_names)
+        for col in ["U1", "U2", "RX", "FC", "US", "UE"]:
+            if col in umi_df.columns:
+                raw_adata.obs[col] = umi_df[col].values
+        del umi_df
+
+    # Load barcode tags from Parquet sidecar (written by extract_and_assign_barcodes_in_bam)
+    if (
+        demux_backend == "smftools"
+        and cfg.barcode_kit
+        and barcode_sidecar
+        and Path(barcode_sidecar).exists()
+    ):
+        logger.info("Loading barcode tags from Parquet sidecar: %s", barcode_sidecar)
+        bc_df = pd.read_parquet(barcode_sidecar).set_index("read_name")
+        bc_df = bc_df.reindex(raw_adata.obs_names)
+        for col in ["BC", "BM", "B1", "B2", "B3", "B4", "B5", "B6"]:
+            if col in bc_df.columns:
+                raw_adata.obs[col] = bc_df[col].values
+        del bc_df
+
     # Expand dorado bi array tag into individual float score columns
     if "bi" in bam_tag_names:
         expand_bi_tag_columns(raw_adata, bi_column="bi")

src/smftools/config/default.yaml

@@ -89,7 +89,6 @@ demux_backend: dorado # smftools|dorado - smftools uses YAML-based barcode refs,
 barcode_both_ends: False # Require barcode match at both ends
 trim: False # dorado adapter and barcode removal during demultiplexing
 use_umi: False # Whether to detect and annotate UMIs in aligned_sorted BAM before demultiplexing
-umi_adapters: [null, null] # Two-slot list [left_ref_end_adapter, right_ref_end_adapter]
 umi_length: null # Length of each UMI (required when use_umi is true)
 umi_search_window: 200 # Max distance from read ends to consider an adapter match for UMI extraction
 umi_adapter_matcher: edlib # exact|edlib