smftools 0.2.3__py3-none-any.whl → 0.2.5__py3-none-any.whl

smftools/cli_entry.py

@@ -1,20 +1,38 @@
+import logging
+from pathlib import Path
+from typing import Sequence
+
 import click
 import pandas as pd
-from pathlib import Path
-from typing import Dict, Optional, Sequence
 
+from .cli.hmm_adata import hmm_adata
 from .cli.load_adata import load_adata
-from .cli.cli_flows import flow_I
 from .cli.preprocess_adata import preprocess_adata
 from .cli.spatial_adata import spatial_adata
-from .cli.hmm_adata import hmm_adata
+from .informatics.pod5_functions import subsample_pod5
+from .logging_utils import setup_logging
+from .readwrite import concatenate_h5ads
 
-from .readwrite import merge_barcoded_anndatas_core, safe_read_h5ad, safe_write_h5ad, concatenate_h5ads
 
 @click.group()
-def cli():
+@click.option(
+    "--log-file",
+    type=click.Path(dir_okay=False, writable=True, path_type=Path),
+    default=None,
+    help="Optional file path to write smftools logs.",
+)
+@click.option(
+    "--log-level",
+    type=click.Choice(["CRITICAL", "ERROR", "WARNING", "INFO", "DEBUG"], case_sensitive=False),
+    default="INFO",
+    show_default=True,
+    help="Logging level for smftools output.",
+)
+def cli(log_file: Path | None, log_level: str):
     """Command-line interface for smftools."""
-    pass
+    level = getattr(logging, log_level.upper(), logging.INFO)
+    setup_logging(level=level, log_file=log_file)
+
 
 ####### Load anndata from raw data ###########
 @cli.command()
@@ -22,32 +40,44 @@ def cli():
 def load(config_path):
     """Load and process data from CONFIG_PATH."""
     load_adata(config_path)
+
+
 ##########################################
 
+
 ####### Preprocessing ###########
 @cli.command()
 @click.argument("config_path", type=click.Path(exists=True))
 def preprocess(config_path):
     """Preprocess data from CONFIG_PATH."""
     preprocess_adata(config_path)
+
+
 ##########################################
 
+
 ####### Spatial ###########
 @cli.command()
 @click.argument("config_path", type=click.Path(exists=True))
 def spatial(config_path):
     """Process data from CONFIG_PATH."""
     spatial_adata(config_path)
+
+
 ##########################################
 
+
 ####### HMM ###########
 @cli.command()
 @click.argument("config_path", type=click.Path(exists=True))
 def hmm(config_path):
     """Process data from CONFIG_PATH."""
     hmm_adata(config_path)
+
+
 ##########################################
 
+
 ####### batch command ###########
 @cli.command()
 @click.argument(
@@ -126,7 +156,9 @@ def batch(task, config_table: Path, column: str, sep: str | None):
                     dtype=str,
                 )
             except Exception as e:
-                raise click.ClickException(f"Failed to read {config_table} as headerless list: {e}") from e
+                raise click.ClickException(
+                    f"Failed to read {config_table} as headerless list: {e}"
+                ) from e
 
             config_series = df[column]
         else:
@@ -137,12 +169,7 @@ def batch(task, config_table: Path, column: str, sep: str | None):
                 )
             config_series = df[column]
 
-        config_paths = (
-            config_series.dropna()
-            .map(str)
-            .map(lambda p: Path(p).expanduser())
-            .tolist()
-        )
+        config_paths = config_series.dropna().map(str).map(lambda p: Path(p).expanduser()).tolist()
 
     # ----------------------------
     # Validate config paths
@@ -163,9 +190,7 @@ def batch(task, config_table: Path, column: str, sep: str | None):
 
     func = task_funcs[task]
 
-    click.echo(
-        f"Running task '{task}' on {len(config_paths)} config paths from {config_table}"
-    )
+    click.echo(f"Running task '{task}' on {len(config_paths)} config paths from {config_table}")
 
     # ----------------------------
     # Loop over paths
@@ -178,13 +203,16 @@ def batch(task, config_table: Path, column: str, sep: str | None):
         click.echo(f"[{i}/{len(config_paths)}] {task} → {cfg}")
 
         try:
-            func(str(cfg))   # underlying functions take a string path
+            func(str(cfg))  # underlying functions take a string path
         except Exception as e:
             click.echo(f"  ERROR on {cfg}: {e}")
 
     click.echo("Batch processing complete.")
+
+
 ##########################################
 
+
 ####### concatenate command ###########
 @cli.command("concatenate")
 @click.argument(
@@ -244,9 +272,9 @@ def concatenate_cmd(
 
     Two modes:
 
-        smftools concatenate out.h5ad --input-dir ./dir
+        smftools concatenate out.h5ad.gz --input-dir ./dir
 
-        smftools concatenate out.h5ad --csv-path paths.csv --csv-column h5ad_path
+        smftools concatenate out.h5ad.gz --csv-path paths.csv --csv-column h5ad_path
 
     TXT input also works (one file path per line).
 
@@ -266,170 +294,57 @@ def concatenate_cmd(
             delete_inputs=delete,
             restore_backups=restore,
         )
-        click.echo(f"✓ Concatenated file written to: {out}")
+        click.echo(f"Concatenated file written to: {out}")
 
     except Exception as e:
         raise click.ClickException(str(e)) from e
+
+
 ##########################################
 
-####### Merging existing anndatas from an experiment that used two different demultiplexing rules #######
-# REQUIRED_KEYS = ("adata_single_path", "adata_double_path")
-# OPTIONAL_KEYS = (
-#     "adata_single_backups_path",
-#     "adata_double_backups_path",
-#     "output_path",
-#     "merged_filename",
-# )
-
-# def _read_config_csv(csv_path: Path) -> Dict[str, str]:
-#     """
-#     Read a multi-row, two-column CSV of key,value pairs into a dict.
-
-#     Supported features:
-#       - Optional header ("key,value") or none.
-#       - Comments starting with '#' and blank lines are ignored.
-#       - If duplicate keys occur, the last one wins.
-#       - Keys are matched literally against REQUIRED_KEYS/OPTIONAL_KEYS.
-#     """
-#     try:
-#         # Read as two columns regardless of header; comments ignored.
-#         df = pd.read_csv(
-#             csv_path,
-#             dtype=str,
-#             comment="#",
-#             header=None,          # treat everything as rows; we'll normalize below
-#             usecols=[0, 1],
-#             names=["key", "value"]
-#         )
-#     except Exception as e:
-#         raise click.ClickException(f"Failed to read CSV: {e}") from e
-
-#     # Drop completely empty rows
-#     df = df.fillna("").astype(str)
-#     df["key"] = df["key"].str.strip()
-#     df["value"] = df["value"].str.strip()
-#     df = df[(df["key"] != "") & (df["key"].notna())]
-
-#     if df.empty:
-#         raise click.ClickException("Config CSV is empty after removing comments/blank lines.")
-
-#     # Remove an optional header row if present
-#     if df.iloc[0]["key"].lower() in {"key", "keys"}:
-#         df = df.iloc[1:]
-#         df = df[(df["key"] != "") & (df["key"].notna())]
-#         if df.empty:
-#             raise click.ClickException("Config CSV contains only a header row.")
-
-#     # Build dict; last occurrence of a key wins
-#     cfg = {}
-#     for k, v in zip(df["key"], df["value"]):
-#         cfg[k] = v
-
-#     # Validate required keys
-#     missing = [k for k in REQUIRED_KEYS if not cfg.get(k)]
-#     if missing:
-#         raise click.ClickException(
-#             "Missing required keys in CSV: "
-#             + ", ".join(missing)
-#             + "\nExpected keys:\n  - "
-#             + "\n  - ".join(REQUIRED_KEYS)
-#             + "\nOptional keys:\n  - "
-#             + "\n  - ".join(OPTIONAL_KEYS)
-#         )
-
-#     return cfg
-
-# def _resolve_output_path(cfg: Dict[str, str], single_path: Path, double_path: Path) -> Path:
-#     """Decide on the output .h5ad path based on CSV; create directories if needed."""
-#     merged_filename = cfg.get("merged_filename") or f"merged_{single_path.stem}__{double_path.stem}.h5ad"
-#     if not merged_filename.endswith(".h5ad"):
-#         merged_filename += ".h5ad"
-
-#     output_path_raw = cfg.get("output_path", "").strip()
-
-#     if not output_path_raw:
-#         out_dir = Path.cwd() / "merged_output"
-#         out_dir.mkdir(parents=True, exist_ok=True)
-#         return out_dir / merged_filename
-
-#     output_path = Path(output_path_raw)
-
-#     if output_path.suffix.lower() == ".h5ad":
-#         output_path.parent.mkdir(parents=True, exist_ok=True)
-#         return output_path
-
-#     # Treat as directory
-#     output_path.mkdir(parents=True, exist_ok=True)
-#     return output_path / merged_filename
-
-# def _maybe_read_adata(label: str, primary: Path, backups: Optional[Path]):
-
-#     if backups:
-#         click.echo(f"Loading {label} from {primary} with backups at {backups} ...")
-#         return safe_read_h5ad(primary, backups_path=backups, restore_backups=True)
-#     else:
-#         click.echo(f"Loading {label} from {primary} with backups disabled ...")
-#         return safe_read_h5ad(primary, restore_backups=False)
-
-
-# @cli.command()
-# @click.argument("config_path", type=click.Path(exists=True, dir_okay=False, readable=True, path_type=Path))
-# def merge_barcoded_anndatas(config_path: Path):
-#     """
-#     Merge two AnnData objects from the same experiment that were demultiplexed
-#     under different end-barcoding requirements, using a 1-row CSV for config.
-
-#     CSV must include:
-#       - adata_single_path
-#       - adata_double_path
-
-#     Optional columns:
-#       - adata_single_backups_path
-#       - adata_double_backups_path
-#       - output_path            (file or directory; default: ./merged_output/)
-#       - merged_filename        (default: merged_<single>__<double>.h5ad)
-
-#     Example CSV:
-
-#         adata_single_path,adata_double_path,adata_single_backups_path,adata_double_backups_path,output_path,merged_filename
-#         /path/single.h5ad,/path/double.h5ad,,,,merged_output,merged_run.h5ad
-#     """
-#     try:
-#         cfg = _read_config_csv(config_path)
-
-#         single_path = Path(cfg["adata_single_path"]).expanduser().resolve()
-#         double_path = Path(cfg["adata_double_path"]).expanduser().resolve()
-
-#         for p, label in [(single_path, "adata_single_path"), (double_path, "adata_double_path")]:
-#             if not p.exists():
-#                 raise click.ClickException(f"{label} does not exist: {p}")
-
-#         single_backups = Path(cfg["adata_single_backups_path"]).expanduser().resolve() if cfg.get("adata_single_backups_path") else None
-#         double_backups = Path(cfg["adata_double_backups_path"]).expanduser().resolve() if cfg.get("adata_double_backups_path") else None
-
-#         if single_backups and not single_backups.exists():
-#             raise click.ClickException(f"adata_single_backups_path does not exist: {single_backups}")
-#         if double_backups and not double_backups.exists():
-#             raise click.ClickException(f"adata_double_backups_path does not exist: {double_backups}")
-
-#         output_path = _resolve_output_path(cfg, single_path, double_path)
-
-#         # Load
-#         adata_single, read_report_single = _maybe_read_adata("single-barcoded AnnData", single_path, single_backups)
-#         adata_double, read_report_double = _maybe_read_adata("double-barcoded AnnData", double_path, double_backups)
-
-#         click.echo("Merging AnnDatas ...")
-#         merged = merge_barcoded_anndatas_core(adata_single, adata_double)
-
-#         click.echo(f"Writing merged AnnData to: {output_path}")
-#         backup_dir = output_path.cwd() / "merged_backups"
-#         safe_write_h5ad(merged, output_path, backup=True, backup_dir=backup_dir)
-
-#         click.secho(f"Done. Merged AnnData saved to {output_path}", fg="green")
-
-#     except click.ClickException:
-#         raise
-#     except Exception as e:
-#         # Surface unexpected errors cleanly
-#         raise click.ClickException(f"Unexpected error: {e}") from e
-################################################################################################################
+
+####### subsample pod5 command ###########
+@cli.command("subsample-pod5")
+@click.argument(
+    "pod5_path",
+    type=click.Path(exists=True, path_type=Path),
+)
+@click.option(
+    "--read-names",
+    "-r",
+    type=click.Path(exists=True, path_type=Path),
+    default=None,
+    help="Text file with one read_id per line.",
+)
+@click.option(
+    "--n-reads",
+    "-n",
+    type=int,
+    default=None,
+    help="Randomly subsample N reads.",
+)
+@click.option(
+    "--outdir",
+    "-o",
+    type=click.Path(path_type=Path, file_okay=False),
+    required=True,
+    help="Output directory for subsampled POD5.",
+)
+def subsample_pod5_cmd(pod5_path, read_names, n_reads, outdir):
+    """
+    Subsample POD5 file(s) by read ID list or random sampling.
+    """
+
+    # --- Validate mutually exclusive options ---
+    if (read_names is None and n_reads is None) or (read_names and n_reads):
+        raise click.UsageError("You must specify exactly ONE of --read-names or --n-reads.")
+
+    outdir.mkdir(parents=True, exist_ok=True)
+
+    subsample_arg = str(read_names) if read_names else n_reads
+
+    subsample_pod5(
+        pod5_path=str(pod5_path),
+        read_name_path=subsample_arg,
+        output_directory=str(outdir),
+    )

smftools/config/__init__.py

@@ -1 +1 @@
-from .experiment_config import LoadExperimentConfig, ExperimentConfig
+from .experiment_config import ExperimentConfig, LoadExperimentConfig

smftools/config/conversion.yaml

@@ -9,6 +9,17 @@ conversion_types:
 # Read QC Params
 read_mod_filtering_use_other_c_as_background: True
 
+# Spatial Analysis - Autocorr params
+autocorr_site_types:
+  - "GpC"
+
+# Spatial Analysis - Clustermap params
+layer_for_clustermap_plotting: 'nan0_0minus1'
+clustermap_cmap_c: "coolwarm"
+clustermap_cmap_gpc: "coolwarm"
+clustermap_cmap_cpg: "viridis"
+clustermap_cmap_a: "coolwarm"
+
 ######## smftools hmm params #########
 # HMM
 cpg: True # whether to use the default HMM endogenous CpG patch params
@@ -18,21 +29,21 @@ hmm_feature_sets:
   footprint:
     state: "Non-Modified"
     features:
-      small_bound_stretch: [10, 30]
-      medium_bound_stretch: [30, 110]
-      putative_nucleosome: [110, 200]
+      small_bound_stretch: [6, 40]
+      medium_bound_stretch: [40, 100]
+      putative_nucleosome: [100, 200]
       large_bound_stretch: [200, inf]
   accessible:
     state: "Modified"
     features:
       small_accessible_patch: [3, 20]
       mid_accessible_patch: [20, 40]
-      mid_large_accessible_patch: [40, 130]
-      large_accessible_patch: [130, inf]
+      large_accessible_patch: [40, 110]
+      nucleosome_depleted_region: [110, inf]
   cpg:
     state: "Modified"
     features:
       cpg_patch: [0, inf]
 
 hmm_merge_layer_features:
-  - ["GpC_all_accessible_features", 80]
+  - ["all_accessible_features", 60]

smftools/config/deaminase.yaml

@@ -7,6 +7,8 @@ conversion_types:
 
 mod_target_bases:
 - "C"
+enzyme_target_bases:
+- "C"
 
 ######## smftools preprocess params #########
 read_mod_filtering_gpc_thresholds:
@@ -15,7 +17,7 @@ read_mod_filtering_gpc_thresholds:
 read_mod_filtering_cpg_thresholds:
   - null
   - null
-read_mod_filtering_any_c_thresholds:
+read_mod_filtering_c_thresholds:
   - 0.01
   - 0.99
 read_mod_filtering_a_thresholds:
@@ -26,16 +28,16 @@ read_mod_filtering_use_other_c_as_background: False
 
 # Duplicate Detection Params
 duplicate_detection_site_types:
-  - "any_C"
+  - "C"
 
 ######## smftools analyze params #########
 # Autocorrelation params
 autocorr_site_types:
-  - "any_C"
+  - "C"
 
 # Correlation matrix params
 correlation_matrix_site_types:
-  - "any_C_site"
+  - "C_site"
 
 # ######## smftools hmm params #########
 cpg: False # whether to use the default HMM endogenous CpG patch params
@@ -45,17 +47,17 @@ hmm_feature_sets:
   footprint:
     state: "Non-Modified"
     features:
-      small_bound_stretch: [10, 30]
-      medium_bound_stretch: [30, 110]
-      putative_nucleosome: [110, 200]
+      small_bound_stretch: [6, 40]
+      medium_bound_stretch: [40, 100]
+      putative_nucleosome: [100, 200]
       large_bound_stretch: [200, inf]
   accessible:
     state: "Modified"
     features:
       small_accessible_patch: [3, 20]
       mid_accessible_patch: [20, 40]
-      mid_large_accessible_patch: [40, 130]
-      large_accessible_patch: [130, inf]
+      large_accessible_patch: [40, 110]
+      nucleosome_depleted_region: [110, inf]
 
 hmm_merge_layer_features:
-  - ["C_all_accessible_features", 80]
+  - ["all_accessible_features", 60]