RNApolis 0.8.0__py3-none-any.whl → 0.8.1__py3-none-any.whl

rnapolis/parser_v2.py

@@ -1,4 +1,6 @@
 import io
+import os
+import string
 import tempfile
 from typing import IO, TextIO, Union
 
@@ -52,23 +54,27 @@ def parse_pdb_atoms(content: Union[str, IO[str]]) -> pd.DataFrame:
             continue
 
         # Parse fields according to PDB format specification
+        alt_loc = line[16:17].strip()
         icode = line[26:27].strip()
+        element = line[76:78].strip()
+        charge = line[78:80].strip()
+
         record = {
             "record_type": record_type,
             "serial": line[6:11].strip(),
             "name": line[12:16].strip(),
-            "altLoc": line[16:17].strip(),
+            "altLoc": None if not alt_loc else alt_loc,  # Store None if empty
             "resName": line[17:20].strip(),
             "chainID": line[21:22].strip(),
             "resSeq": line[22:26].strip(),
-            "iCode": None if not icode else icode,  # Convert empty string to None
+            "iCode": None if not icode else icode,  # Store None if empty
             "x": line[30:38].strip(),
             "y": line[38:46].strip(),
             "z": line[46:54].strip(),
             "occupancy": line[54:60].strip(),
             "tempFactor": line[60:66].strip(),
-            "element": line[76:78].strip(),
-            "charge": line[78:80].strip(),
+            "element": None if not element else element,  # Store None if empty
+            "charge": None if not charge else charge,  # Store None if empty
             "model": current_model,  # Add the current model number
         }
 
@@ -149,18 +155,37 @@ def parse_cif_atoms(content: Union[str, IO[str]]) -> pd.DataFrame:
     """
     adapter = IoAdapterPy()
 
-    # Handle both string content and file-like objects
+    # Handle string, StringIO, and file-like objects
     if isinstance(content, str):
-        # Create a temporary file to use with the adapter
-        import tempfile
-
-        with tempfile.NamedTemporaryFile(mode="w+", suffix=".cif") as temp_file:
+        # Create a temporary file for string input
+        with tempfile.NamedTemporaryFile(
+            mode="w+", suffix=".cif", delete=False
+        ) as temp_file:
             temp_file.write(content)
-            temp_file.flush()
-            data = adapter.readFile(temp_file.name)
-    else:
-        # Assume it's a file-like object with a name attribute
+            temp_file_path = temp_file.name
+        try:
+            data = adapter.readFile(temp_file_path)
+        finally:
+            os.remove(temp_file_path)  # Clean up the temporary file
+    elif isinstance(content, io.StringIO):
+        # Create a temporary file for StringIO input
+        with tempfile.NamedTemporaryFile(
+            mode="w+", suffix=".cif", delete=False
+        ) as temp_file:
+            content.seek(0)  # Ensure reading from the start
+            temp_file.write(content.read())
+            temp_file_path = temp_file.name
+        try:
+            data = adapter.readFile(temp_file_path)
+        finally:
+            os.remove(temp_file_path)  # Clean up the temporary file
+    elif hasattr(content, "name"):
+        # Assume it's a file-like object with a name attribute (like an open file)
         data = adapter.readFile(content.name)
+    else:
+        raise TypeError(
+            "Unsupported input type for parse_cif_atoms. Expected str, file-like object with name, or StringIO."
+        )
 
     # Get the atom_site category
     category = data[0].getObj("atom_site")
@@ -176,47 +201,133 @@ def parse_cif_atoms(content: Union[str, IO[str]]) -> pd.DataFrame:
     # Create a list of dictionaries for each atom
     records = []
     for row in rows:
-        record = dict(zip(attributes, row))
-
-        # Convert "?" or "." in insertion code to None
-        if "pdbx_PDB_ins_code" in record:
-            if record["pdbx_PDB_ins_code"] in ["?", ".", ""]:
-                record["pdbx_PDB_ins_code"] = None
-
+        record = {}
+        for attr, value in zip(attributes, row):
+            # Store None if value indicates missing data ('?' or '.')
+            if value in ["?", "."]:
+                record[attr] = None
+            else:
+                record[attr] = value
         records.append(record)
 
     # Create DataFrame from records
     df = pd.DataFrame(records)
 
-    # Convert numeric columns to appropriate types
-    numeric_columns = [
-        "id",
-        "auth_seq_id",
+    # Define columns based on mmCIF specification for atom_site
+    float_cols = [
+        "aniso_B[1][1]",
+        "aniso_B[1][1]_esd",
+        "aniso_B[1][2]",
+        "aniso_B[1][2]_esd",
+        "aniso_B[1][3]",
+        "aniso_B[1][3]_esd",
+        "aniso_B[2][2]",
+        "aniso_B[2][2]_esd",
+        "aniso_B[2][3]",
+        "aniso_B[2][3]_esd",
+        "aniso_B[3][3]",
+        "aniso_B[3][3]_esd",
+        "aniso_ratio",
+        "aniso_U[1][1]",
+        "aniso_U[1][1]_esd",
+        "aniso_U[1][2]",
+        "aniso_U[1][2]_esd",
+        "aniso_U[1][3]",
+        "aniso_U[1][3]_esd",
+        "aniso_U[2][2]",
+        "aniso_U[2][2]_esd",
+        "aniso_U[2][3]",
+        "aniso_U[2][3]_esd",
+        "aniso_U[3][3]",
+        "aniso_U[3][3]_esd",
+        "B_equiv_geom_mean",
+        "B_equiv_geom_mean_esd",
+        "B_iso_or_equiv",
+        "B_iso_or_equiv_esd",
         "Cartn_x",
+        "Cartn_x_esd",
         "Cartn_y",
+        "Cartn_y_esd",
         "Cartn_z",
+        "Cartn_z_esd",
+        "fract_x",
+        "fract_x_esd",
+        "fract_y",
+        "fract_y_esd",
+        "fract_z",
+        "fract_z_esd",
         "occupancy",
-        "B_iso_or_equiv",
+        "occupancy_esd",
+        "U_equiv_geom_mean",
+        "U_equiv_geom_mean_esd",
+        "U_iso_or_equiv",
+        "U_iso_or_equiv_esd",
+    ]
+    int_cols = [
+        "attached_hydrogens",
+        "label_seq_id",
+        "symmetry_multiplicity",
+        "pdbx_PDB_model_num",
         "pdbx_formal_charge",
+        "pdbx_label_index",
     ]
-
-    for col in numeric_columns:
-        if col in df.columns:
-            df[col] = pd.to_numeric(df[col], errors="coerce")
-
-    # Convert categorical columns
-    categorical_columns = [
+    category_cols = [
+        "auth_asym_id",
+        "auth_atom_id",
+        "auth_comp_id",
+        "auth_seq_id",
+        "calc_attached_atom",
+        "calc_flag",
+        "disorder_assembly",
+        "disorder_group",
         "group_PDB",
-        "type_symbol",
+        "id",
+        "label_alt_id",
+        "label_asym_id",
         "label_atom_id",
         "label_comp_id",
-        "label_asym_id",
-        "auth_atom_id",
-        "auth_comp_id",
-        "auth_asym_id",
+        "label_entity_id",
+        "thermal_displace_type",
+        "type_symbol",
+        "pdbx_atom_ambiguity",
+        "adp_type",
+        "refinement_flags",
+        "refinement_flags_adp",
+        "refinement_flags_occupancy",
+        "refinement_flags_posn",
+        "pdbx_auth_alt_id",
+        "pdbx_PDB_ins_code",
+        "pdbx_PDB_residue_no",
+        "pdbx_PDB_residue_name",
+        "pdbx_PDB_strand_id",
+        "pdbx_PDB_atom_name",
+        "pdbx_auth_atom_name",
+        "pdbx_auth_comp_id",
+        "pdbx_auth_asym_id",
+        "pdbx_auth_seq_id",
+        "pdbx_tls_group_id",
+        "pdbx_ncs_dom_id",
+        "pdbx_group_NDB",
+        "pdbx_atom_group",
+        "pdbx_label_seq_num",
+        "pdbx_not_in_asym",
+        "pdbx_sifts_xref_db_name",
+        "pdbx_sifts_xref_db_acc",
+        "pdbx_sifts_xref_db_num",
+        "pdbx_sifts_xref_db_res",
     ]
 
-    for col in categorical_columns:
+    # Convert columns to appropriate types
+    for col in float_cols:
+        if col in df.columns:
+            df[col] = pd.to_numeric(df[col], errors="coerce")
+
+    for col in int_cols:
+        if col in df.columns:
+            # Use Int64 (nullable integer) to handle potential NaNs from coercion
+            df[col] = pd.to_numeric(df[col], errors="coerce").astype("Int64")
+
+    for col in category_cols:
         if col in df.columns:
             df[col] = df[col].astype("category")
 
@@ -226,6 +337,473 @@ def parse_cif_atoms(content: Union[str, IO[str]]) -> pd.DataFrame:
     return df
 
 
+def can_write_pdb(df: pd.DataFrame) -> bool:
+    """
+    Check if the DataFrame can be losslessly represented in PDB format.
+
+    PDB format has limitations on field widths:
+    - Atom serial number (id): max 99999
+    - Chain identifier (auth_asym_id): max 1 character
+    - Residue sequence number (auth_seq_id): max 9999
+
+    Parameters:
+    -----------
+    df : pd.DataFrame
+        DataFrame containing atom records, as created by parse_pdb_atoms or parse_cif_atoms.
+
+    Returns:
+    --------
+    bool
+        True if the DataFrame can be written to PDB format without data loss/truncation, False otherwise.
+    """
+    format_type = df.attrs.get("format")
+
+    if format_type == "PDB":
+        # Assume data originally from PDB already fits PDB constraints
+        return True
+
+    if df.empty:
+        # An empty DataFrame can be represented as an empty PDB file
+        return True
+
+    if format_type == "mmCIF":
+        # Check serial number (id)
+        # Convert to numeric first to handle potential categorical type and NaNs
+        if "id" not in df.columns or (
+            pd.to_numeric(df["id"], errors="coerce").max() > 99999
+        ):
+            return False
+
+        # Check chain ID (auth_asym_id) length
+        if "auth_asym_id" not in df.columns or (
+            df["auth_asym_id"].dropna().astype(str).str.len().max() > 1
+        ):
+            return False
+
+        # Check residue sequence number (auth_seq_id)
+        if "auth_seq_id" not in df.columns or (
+            pd.to_numeric(df["auth_seq_id"], errors="coerce").max() > 9999
+        ):
+            return False
+
+        # All checks passed for mmCIF
+        return True
+
+    # If format is unknown or not PDB/mmCIF, assume it cannot be safely written
+    return False
+
+
+def fit_to_pdb(df: pd.DataFrame) -> pd.DataFrame:
+    """
+    Attempts to fit the atom data in a DataFrame to comply with PDB format limitations.
+
+    If the data already fits (checked by can_write_pdb), returns the original DataFrame.
+    Otherwise, checks if fitting is possible based on total atoms, unique chains,
+    and residues per chain. If fitting is possible, it renumbers atoms, renames chains,
+    and renumbers residues within each chain sequentially starting from 1.
+
+    Parameters:
+    -----------
+    df : pd.DataFrame
+        DataFrame containing atom records, as created by parse_pdb_atoms or parse_cif_atoms.
+
+    Returns:
+    --------
+    pd.DataFrame
+        A new DataFrame with data potentially modified to fit PDB constraints.
+        The 'format' attribute of the returned DataFrame will be set to 'PDB'.
+
+    Raises:
+    -------
+    ValueError
+        If the data cannot be fitted into PDB format constraints (too many atoms,
+        chains, or residues per chain).
+    """
+    format_type = df.attrs.get("format")
+
+    if not format_type:
+        raise ValueError("DataFrame format attribute is not set.")
+
+    if can_write_pdb(df):
+        return df
+
+    # Determine column names based on format
+    if format_type == "PDB":
+        serial_col = "serial"
+        chain_col = "chainID"
+        resseq_col = "resSeq"
+        icode_col = "iCode"
+    elif format_type == "mmCIF":
+        serial_col = "id"
+        chain_col = "auth_asym_id"
+        resseq_col = "auth_seq_id"
+        icode_col = "pdbx_PDB_ins_code"
+    else:
+        raise ValueError(f"Unsupported DataFrame format: {format_type}")
+
+    # --- Feasibility Checks ---
+    if chain_col not in df.columns:
+        raise ValueError(f"Missing required chain column: {chain_col}")
+    if resseq_col not in df.columns:
+        raise ValueError(f"Missing required residue sequence column: {resseq_col}")
+
+    unique_chains = df[chain_col].unique()
+    num_chains = len(unique_chains)
+    total_atoms = len(df)
+    max_pdb_serial = 99999
+    max_pdb_residue = 9999
+    available_chain_ids = list(
+        string.ascii_uppercase + string.ascii_lowercase + string.digits
+    )
+    max_pdb_chains = len(available_chain_ids)
+
+    # Check 1: Total atoms + TER lines <= 99999
+    if total_atoms + num_chains > max_pdb_serial:
+        raise ValueError(
+            f"Cannot fit to PDB: Total atoms ({total_atoms}) + TER lines ({num_chains}) exceeds PDB limit ({max_pdb_serial})."
+        )
+
+    # Check 2: Number of chains <= 62
+    if num_chains > max_pdb_chains:
+        raise ValueError(
+            f"Cannot fit to PDB: Number of unique chains ({num_chains}) exceeds PDB limit ({max_pdb_chains})."
+        )
+
+    # Check 3: Max residues per chain <= 9999
+    # More accurate check: group by chain, then count unique (resSeq, iCode) tuples
+    # Use a temporary structure to avoid modifying the original df
+    check_df = pd.DataFrame(
+        {
+            "chain": df[chain_col],
+            "resSeq": df[resseq_col],
+            "iCode": df[icode_col].fillna("") if icode_col in df.columns else "",
+        }
+    )
+    residue_counts = check_df.groupby("chain").apply(
+        lambda x: x[["resSeq", "iCode"]].drop_duplicates().shape[0]
+    )
+    max_residues_per_chain = residue_counts.max() if not residue_counts.empty else 0
+
+    if max_residues_per_chain > max_pdb_residue:
+        raise ValueError(
+            f"Cannot fit to PDB: Maximum residues in a single chain ({max_residues_per_chain}) exceeds PDB limit ({max_pdb_residue})."
+        )
+
+    # --- Perform Fitting ---
+    df_fitted = df.copy()
+
+    # 1. Rename Chains
+    chain_mapping = {
+        orig_chain: available_chain_ids[i] for i, orig_chain in enumerate(unique_chains)
+    }
+    df_fitted[chain_col] = df_fitted[chain_col].map(chain_mapping)
+    # Ensure the chain column is treated as string/object after mapping
+    df_fitted[chain_col] = df_fitted[chain_col].astype(object)
+
+    # 2. Renumber Residues within each new chain
+    new_resseq_col = "new_resSeq"  # Temporary column for new numbering
+    df_fitted[new_resseq_col] = -1  # Initialize
+
+    all_new_res_maps = {}
+    for new_chain_id, group in df_fitted.groupby(chain_col):
+        # Identify unique original residues (seq + icode) in order of appearance
+        original_residues = group[[resseq_col, icode_col]].drop_duplicates()
+        # Create mapping: (orig_resSeq, orig_iCode) -> new_resSeq (1-based)
+        residue_mapping = {
+            tuple(res): i + 1
+            for i, res in enumerate(original_residues.itertuples(index=False))
+        }
+        all_new_res_maps[new_chain_id] = residue_mapping
+
+        # Apply mapping to the group
+        res_indices = group.set_index([resseq_col, icode_col]).index
+        df_fitted.loc[group.index, new_resseq_col] = res_indices.map(residue_mapping)
+
+    # Replace original residue number and clear insertion code
+    df_fitted[resseq_col] = df_fitted[new_resseq_col]
+    df_fitted[icode_col] = None  # Insertion codes are now redundant
+    df_fitted.drop(columns=[new_resseq_col], inplace=True)
+    # Convert resseq_col back to Int64 if it was before, handling potential NaNs if any step failed
+    df_fitted[resseq_col] = df_fitted[resseq_col].astype("Int64")
+
+    # 3. Renumber Atom Serials
+    new_serial_col = "new_serial"
+    df_fitted[new_serial_col] = -1  # Initialize
+    current_serial = 0
+    last_chain_id_for_serial = None
+
+    # Iterate in the potentially re-sorted order after grouping/mapping
+    # Ensure stable sort order for consistent serial numbering
+    df_fitted.sort_index(
+        inplace=True
+    )  # Sort by original index to maintain original atom order as much as possible
+
+    for index, row in df_fitted.iterrows():
+        current_chain_id = row[chain_col]
+        if (
+            last_chain_id_for_serial is not None
+            and current_chain_id != last_chain_id_for_serial
+        ):
+            current_serial += 1  # Increment for TER line
+
+        current_serial += 1
+        if current_serial > max_pdb_serial:
+            # This should have been caught by the initial check, but is a safeguard
+            raise ValueError("Serial number exceeded PDB limit during renumbering.")
+
+        df_fitted.loc[index, new_serial_col] = current_serial
+        last_chain_id_for_serial = current_chain_id
+
+    # Replace original serial number
+    df_fitted[serial_col] = df_fitted[new_serial_col]
+    df_fitted.drop(columns=[new_serial_col], inplace=True)
+    # Convert serial_col back to Int64
+    df_fitted[serial_col] = df_fitted[serial_col].astype("Int64")
+
+    # Update attributes and column types for PDB compatibility
+    df_fitted.attrs["format"] = "PDB"
+
+    # Ensure final column types match expected PDB output (especially categories)
+    # Reapply categorical conversion as some operations might change dtypes
+    pdb_categorical_cols = [
+        "record_type",
+        "name",
+        "altLoc",
+        "resName",
+        chain_col,
+        "element",
+        "charge",
+        icode_col,
+    ]
+    if "record_type" not in df_fitted.columns and "group_PDB" in df_fitted.columns:
+        df_fitted.rename(
+            columns={"group_PDB": "record_type"}, inplace=True
+        )  # Ensure correct name
+
+    for col in pdb_categorical_cols:
+        if col in df_fitted.columns:
+            # Handle None explicitly before converting to category if needed
+            if df_fitted[col].isnull().any():
+                df_fitted[col] = (
+                    df_fitted[col].astype(object).fillna("")
+                )  # Fill None with empty string for category
+            df_fitted[col] = df_fitted[col].astype("category")
+
+    # Rename columns if necessary from mmCIF to PDB standard names
+    rename_map = {
+        "id": "serial",
+        "auth_asym_id": "chainID",
+        "auth_seq_id": "resSeq",
+        "pdbx_PDB_ins_code": "iCode",
+        "label_atom_id": "name",  # Prefer label_atom_id if auth_atom_id not present? PDB uses 'name'
+        "label_comp_id": "resName",  # Prefer label_comp_id if auth_comp_id not present? PDB uses 'resName'
+        "type_symbol": "element",
+        "pdbx_formal_charge": "charge",
+        "Cartn_x": "x",
+        "Cartn_y": "y",
+        "Cartn_z": "z",
+        "B_iso_or_equiv": "tempFactor",
+        "group_PDB": "record_type",
+        "pdbx_PDB_model_num": "model",
+        # Add mappings for auth_atom_id -> name, auth_comp_id -> resName if needed,
+        # deciding on precedence if both label_* and auth_* exist.
+        # Current write_pdb prioritizes auth_* when reading mmCIF, so map those.
+        "auth_atom_id": "name",
+        "auth_comp_id": "resName",
+    }
+
+    # Only rename columns that actually exist in the DataFrame
+    actual_rename_map = {k: v for k, v in rename_map.items() if k in df_fitted.columns}
+    df_fitted.rename(columns=actual_rename_map, inplace=True)
+
+    # Ensure essential PDB columns exist, even if empty, if they were created during fitting
+    pdb_essential_cols = [
+        "record_type",
+        "serial",
+        "name",
+        "altLoc",
+        "resName",
+        "chainID",
+        "resSeq",
+        "iCode",
+        "x",
+        "y",
+        "z",
+        "occupancy",
+        "tempFactor",
+        "element",
+        "charge",
+        "model",
+    ]
+    for col in pdb_essential_cols:
+        if col not in df_fitted.columns:
+            # This case might occur if input mmCIF was missing fundamental columns mapped to PDB essentials
+            # Decide on default value or raise error. Adding empty series for now.
+            df_fitted[col] = pd.Series(
+                dtype="object"
+            )  # Add as object to handle potential None/mixed types initially
+
+    # Re-order columns to standard PDB order for clarity
+    final_pdb_order = [col for col in pdb_essential_cols if col in df_fitted.columns]
+    other_cols = [col for col in df_fitted.columns if col not in final_pdb_order]
+    df_fitted = df_fitted[final_pdb_order + other_cols]
+
+    # --- Final Type Conversions for PDB format ---
+    # Convert numeric columns (similar to parse_pdb_atoms)
+    pdb_numeric_columns = [
+        "serial",
+        "resSeq",
+        "x",
+        "y",
+        "z",
+        "occupancy",
+        "tempFactor",
+        "model",
+    ]
+    for col in pdb_numeric_columns:
+        if col in df_fitted.columns:
+            # Use Int64 for integer-like columns that might have been NaN during processing
+            if col in ["serial", "resSeq", "model"]:
+                df_fitted[col] = pd.to_numeric(df_fitted[col], errors="coerce").astype(
+                    "Int64"
+                )
+            else:  # Floats
+                df_fitted[col] = pd.to_numeric(df_fitted[col], errors="coerce")
+
+    # Convert categorical columns (similar to parse_pdb_atoms)
+    # Note: chainID and iCode were already handled during fitting/renaming
+    pdb_categorical_columns_final = [
+        "record_type",
+        "name",
+        "altLoc",
+        "resName",
+        "chainID",  # Already category, but ensure consistency
+        "iCode",  # Already category, but ensure consistency
+        "element",
+        "charge",
+    ]
+    for col in pdb_categorical_columns_final:
+        if col in df_fitted.columns:
+            # Ensure the column is categorical first
+            if not pd.api.types.is_categorical_dtype(df_fitted[col]):
+                # Convert non-categorical columns, handling potential NaNs
+                if df_fitted[col].isnull().any():
+                    df_fitted[col] = (
+                        df_fitted[col].astype(object).fillna("").astype("category")
+                    )
+                else:
+                    df_fitted[col] = df_fitted[col].astype("category")
+            else:
+                # If already categorical, check if '' needs to be added before fillna
+                has_nans = df_fitted[col].isnull().any()
+                if has_nans and "" not in df_fitted[col].cat.categories:
+                    # Add '' category explicitly
+                    df_fitted[col] = df_fitted[col].cat.add_categories([""])
+
+                # Fill None/NaN with empty string (now safe)
+                if has_nans:
+                    df_fitted[col].fillna("", inplace=True)
+
+    return df_fitted
+
+
+def _format_pdb_atom_line(atom_data: dict) -> str:
+    """Formats a dictionary of atom data into a PDB ATOM/HETATM line."""
+    # PDB format specification:
+    # COLUMNS        DATA TYPE     FIELD         DEFINITION
+    # -----------------------------------------------------------------------
+    #  1 -  6        Record name   "ATOM  " or "HETATM"
+    #  7 - 11        Integer       serial        Atom serial number.
+    # 13 - 16        Atom          name          Atom name.
+    # 17             Character     altLoc        Alternate location indicator.
+    # 18 - 20        Residue name  resName       Residue name.
+    # 22             Character     chainID       Chain identifier.
+    # 23 - 26        Integer       resSeq        Residue sequence number.
+    # 27             AChar         iCode         Code for insertion of residues.
+    # 31 - 38        Real(8.3)     x             Orthogonal coordinates for X.
+    # 39 - 46        Real(8.3)     y             Orthogonal coordinates for Y.
+    # 47 - 54        Real(8.3)     z             Orthogonal coordinates for Z.
+    # 55 - 60        Real(6.2)     occupancy     Occupancy.
+    # 61 - 66        Real(6.2)     tempFactor    Temperature factor.
+    # 77 - 78        LString(2)    element       Element symbol, right-justified.
+    # 79 - 80        LString(2)    charge        Charge on the atom.
+
+    # Record name (ATOM/HETATM)
+    record_name = atom_data.get("record_name", "ATOM").ljust(6)
+
+    # Serial number
+    serial = str(atom_data.get("serial", 0)).rjust(5)
+
+    # Atom name - special alignment rules
+    atom_name = atom_data.get("name", "")
+    if len(atom_name) < 4 and atom_name[:1].isalpha():
+        # Pad with space on left for 1-3 char names starting with a letter
+        atom_name_fmt = (" " + atom_name).ljust(4)
+    else:
+        # Use as is, left-justified, for 4-char names or those starting with a digit
+        atom_name_fmt = atom_name.ljust(4)
+
+    # Alternate location indicator
+    alt_loc = atom_data.get("altLoc", "")[:1].ljust(1)  # Max 1 char
+
+    # Residue name
+    res_name = atom_data.get("resName", "").rjust(
+        3
+    )  # Spec says "Residue name", examples often right-justified
+
+    # Chain identifier
+    chain_id = atom_data.get("chainID", "")[:1].ljust(1)  # Max 1 char
+
+    # Residue sequence number
+    res_seq = str(atom_data.get("resSeq", 0)).rjust(4)
+
+    # Insertion code
+    icode = atom_data.get("iCode", "")[:1].ljust(1)  # Max 1 char
+
+    # Coordinates
+    x = f"{atom_data.get('x', 0.0):8.3f}"
+    y = f"{atom_data.get('y', 0.0):8.3f}"
+    z = f"{atom_data.get('z', 0.0):8.3f}"
+
+    # Occupancy
+    occupancy = f"{atom_data.get('occupancy', 1.0):6.2f}"
+
+    # Temperature factor
+    temp_factor = f"{atom_data.get('tempFactor', 0.0):6.2f}"
+
+    # Element symbol
+    element = atom_data.get("element", "").rjust(2)
+
+    # Charge
+    charge_val = atom_data.get("charge", "")
+    charge_fmt = ""
+    if charge_val:
+        try:
+            # Try converting numeric charge (e.g., +1, -2) to PDB format (1+, 2-)
+            charge_int = int(float(charge_val))  # Use float first for cases like "1.0"
+            if charge_int != 0:
+                charge_fmt = f"{abs(charge_int)}{'+' if charge_int > 0 else '-'}"
+        except ValueError:
+            # If already formatted (e.g., "1+", "FE2+"), use its string representation
+            charge_fmt = str(charge_val)
+        # Ensure it fits and is right-justified
+        charge_fmt = charge_fmt.strip()[:2].rjust(2)
+    else:
+        charge_fmt = "  "  # Blank if no charge
+
+    # Construct the full line
+    # Ensure spacing is correct according to the spec
+    # 1-6 Record name | 7-11 Serial | 12 Space | 13-16 Name | 17 AltLoc | 18-20 ResName | 21 Space | 22 ChainID | 23-26 ResSeq | 27 iCode | 28-30 Spaces | 31-38 X | 39-46 Y | 47-54 Z | 55-60 Occupancy | 61-66 TempFactor | 67-76 Spaces | 77-78 Element | 79-80 Charge
+    line = (
+        f"{record_name}{serial} {atom_name_fmt}{alt_loc}{res_name} {chain_id}{res_seq}{icode}   "
+        f"{x}{y}{z}{occupancy}{temp_factor}          "  # 10 spaces
+        f"{element}{charge_fmt}"
+    )
+
+    # Ensure the line is exactly 80 characters long
+    return line.ljust(80)
+
+
 def write_pdb(
     df: pd.DataFrame, output: Union[str, TextIO, None] = None
 ) -> Union[str, None]:
@@ -235,7 +813,8 @@ def write_pdb(
     Parameters:
     -----------
     df : pd.DataFrame
-        DataFrame containing atom records, as created by parse_pdb_atoms or parse_cif_atoms
+        DataFrame containing atom records, as created by parse_pdb_atoms or parse_cif_atoms.
+        Must contain columns mappable to PDB format fields.
     output : Union[str, TextIO, None], optional
         Output file path or file-like object. If None, returns the PDB content as a string.
 
@@ -244,218 +823,162 @@ def write_pdb(
     Union[str, None]
         If output is None, returns the PDB content as a string. Otherwise, returns None.
     """
-    # Create a buffer to store the PDB content
     buffer = io.StringIO()
+    format_type = df.attrs.get("format", "PDB")  # Assume PDB if not specified
 
-    # Get the format of the DataFrame
-    format_type = df.attrs.get("format", "PDB")
-
-    # Variables to track chain changes for TER records
+    last_model_num = None
     last_chain_id = None
-    last_res_seq = None
-    last_res_name = None
-    last_serial = None
-    last_icode = None
-
-    # Process each row in the DataFrame
-    for index, row in df.iterrows():
-        # Get current chain ID
+    last_res_info = None  # Tuple (resSeq, iCode, resName) for TER record
+    last_serial = 0
+
+    # Check if DataFrame is empty
+    if df.empty:
+        buffer.write("END\n")
+        content = buffer.getvalue()
+        buffer.close()
+        if output is not None:
+            if isinstance(output, str):
+                with open(output, "w") as f:
+                    f.write(content)
+            else:
+                output.write(content)
+            return None
+        return content
+
+    for _, row in df.iterrows():
+        atom_data = {}
+
+        # --- Data Extraction ---
         if format_type == "PDB":
-            current_chain_id = row["chainID"]
-        else:  # mmCIF
-            current_chain_id = row.get("auth_asym_id", row.get("label_asym_id", ""))
+            # Pre-process PDB values, converting None to empty strings for optional fields
+            raw_alt_loc = row.get("altLoc")
+            pdb_alt_loc = "" if pd.isna(raw_alt_loc) else str(raw_alt_loc)
+
+            raw_icode = row.get("iCode")
+            pdb_icode = "" if pd.isna(raw_icode) else str(raw_icode)
+
+            raw_element = row.get("element")
+            pdb_element = "" if pd.isna(raw_element) else str(raw_element)
+
+            raw_charge = row.get("charge")
+            pdb_charge = "" if pd.isna(raw_charge) else str(raw_charge)
+
+            atom_data = {
+                "record_name": row.get("record_type", "ATOM"),
+                "serial": int(row.get("serial", 0)),
+                "name": str(row.get("name", "")),
+                "altLoc": pdb_alt_loc,
+                "resName": str(row.get("resName", "")),
+                "chainID": str(row.get("chainID", "")),
+                "resSeq": int(row.get("resSeq", 0)),
+                "iCode": pdb_icode,
+                "x": float(row.get("x", 0.0)),
+                "y": float(row.get("y", 0.0)),
+                "z": float(row.get("z", 0.0)),
+                "occupancy": float(row.get("occupancy", 1.0)),
+                "tempFactor": float(row.get("tempFactor", 0.0)),
+                "element": pdb_element,
+                "charge": pdb_charge,
+                "model": int(row.get("model", 1)),
+            }
+        elif format_type == "mmCIF":
+            # Pre-process mmCIF values to PDB compatible format, converting None to empty strings
+            raw_alt_loc = row.get("label_alt_id")
+            pdb_alt_loc = "" if pd.isna(raw_alt_loc) else str(raw_alt_loc)
+
+            raw_icode = row.get("pdbx_PDB_ins_code")
+            pdb_icode = "" if pd.isna(raw_icode) else str(raw_icode)
+
+            raw_element = row.get("type_symbol")
+            pdb_element = "" if pd.isna(raw_element) else str(raw_element)
+
+            raw_charge = row.get("pdbx_formal_charge")
+            pdb_charge = "" if pd.isna(raw_charge) else str(raw_charge)
+
+            atom_data = {
+                "record_name": row.get("group_PDB", "ATOM"),
+                "serial": int(row.get("id", 0)),
+                "name": str(row.get("auth_atom_id", row.get("label_atom_id", ""))),
+                "altLoc": pdb_alt_loc,
+                "resName": str(row.get("auth_comp_id", row.get("label_comp_id", ""))),
+                "chainID": str(row.get("auth_asym_id", row.get("label_asym_id"))),
+                "resSeq": int(row.get("auth_seq_id", row.get("label_seq_id", 0))),
+                "iCode": pdb_icode,
+                "x": float(row.get("Cartn_x", 0.0)),
+                "y": float(row.get("Cartn_y", 0.0)),
+                "z": float(row.get("Cartn_z", 0.0)),
+                "occupancy": float(row.get("occupancy", 1.0)),
+                "tempFactor": float(row.get("B_iso_or_equiv", 0.0)),
+                "element": pdb_element,
+                "charge": pdb_charge,
+                "model": int(row.get("pdbx_PDB_model_num", 1)),
+            }
+        else:
+            raise ValueError(f"Unsupported DataFrame format: {format_type}")
+
+        # --- MODEL/ENDMDL Records ---
+        current_model_num = atom_data["model"]
+        if current_model_num != last_model_num:
+            if last_model_num is not None:
+                buffer.write("ENDMDL\n")
+            buffer.write(f"MODEL     {current_model_num:>4}\n")
+            last_model_num = current_model_num
+            # Reset chain/residue tracking for the new model
+            last_chain_id = None
+            last_res_info = None
+
+        # --- TER Records ---
+        current_chain_id = atom_data["chainID"]
+        current_res_info = (
+            atom_data["resSeq"],
+            atom_data["iCode"],
+            atom_data["resName"],
+        )
 
-        # Write TER record if chain changes
+        # Write TER if chain ID changes within the same model
         if last_chain_id is not None and current_chain_id != last_chain_id:
-            # Format TER record according to PDB specification
-            # Columns:
-            # 1-6: "TER   "
-            # 7-11: Serial number (right-justified)
-            # 18-20: Residue name (right-justified)
-            # 22: Chain ID
-            # 23-26: Residue sequence number (right-justified)
-            # 27: Insertion code
             ter_serial = str(last_serial + 1).rjust(5)
-            ter_res_name = last_res_name.strip().ljust(3)  # Strip and left-justify
+            ter_res_name = last_res_info[2].strip().rjust(3)  # Use last residue's name
             ter_chain_id = last_chain_id
-            ter_res_seq = last_res_seq.rjust(4)
-            ter_icode = last_icode if last_icode else ""  # Use last recorded iCode
+            ter_res_seq = str(last_res_info[0]).rjust(4)  # Use last residue's seq num
+            ter_icode = (
+                last_res_info[1] if last_res_info[1] else ""
+            )  # Use last residue's icode
 
-            # Construct the TER line ensuring correct spacing for all fields
-            # TER (1-6), serial (7-11), space (12-17), resName (18-20), space (21),
-            # chainID (22), resSeq (23-26), iCode (27)
             ter_line = f"TER   {ter_serial}      {ter_res_name} {ter_chain_id}{ter_res_seq}{ter_icode}"
             buffer.write(ter_line.ljust(80) + "\n")
 
-        # Initialize the line with spaces
-        line = " " * 80
-
-        # Set record type (ATOM or HETATM)
-        if format_type == "PDB":
-            record_type = row["record_type"]
-        else:  # mmCIF
-            record_type = row.get("group_PDB", "ATOM")
-        line = record_type.ljust(6) + line[6:]
-
-        # Set atom serial number
-        if format_type == "PDB":
-            serial = str(int(row["serial"]))
-        else:  # mmCIF
-            serial = str(int(row["id"]))
-        line = line[:6] + serial.rjust(5) + line[11:]
-
-        # Set atom name
-        if format_type == "PDB":
-            atom_name = row["name"]
-        else:  # mmCIF
-            atom_name = row.get("auth_atom_id", row.get("label_atom_id", ""))
-
-        # Right-justify atom name if it starts with a number
-        if atom_name and atom_name[0].isdigit():
-            line = line[:12] + atom_name.ljust(4) + line[16:]
-        else:
-            line = line[:12] + " " + atom_name.ljust(3) + line[16:]
-
-        # Set alternate location indicator
-        if format_type == "PDB":
-            alt_loc = row.get("altLoc", "")
-        else:  # mmCIF
-            alt_loc = row.get("label_alt_id", "")
-        line = line[:16] + alt_loc + line[17:]
-
-        # Set residue name
-        if format_type == "PDB":
-            res_name = row["resName"]
-        else:  # mmCIF
-            res_name = row.get("auth_comp_id", row.get("label_comp_id", ""))
-        line = line[:17] + res_name.ljust(3) + line[20:]
-
-        # Set chain identifier
-        if format_type == "PDB":
-            chain_id = row["chainID"]
-        else:  # mmCIF
-            chain_id = row.get("auth_asym_id", row.get("label_asym_id", ""))
-        line = line[:21] + chain_id + line[22:]
-
-        # Set residue sequence number
-        if format_type == "PDB":
-            res_seq = str(int(row["resSeq"]))
-        else:  # mmCIF
-            res_seq = str(int(row.get("auth_seq_id", row.get("label_seq_id", 0))))
-        line = line[:22] + res_seq.rjust(4) + line[26:]
-
-        # Set insertion code
-        if format_type == "PDB":
-            icode = row["iCode"] if pd.notna(row["iCode"]) else ""
-        else:  # mmCIF
-            icode = (
-                row.get("pdbx_PDB_ins_code", "")
-                if pd.notna(row.get("pdbx_PDB_ins_code", ""))
-                else ""
-            )
-        line = line[:26] + icode + line[27:]
-
-        # Set X coordinate
-        if format_type == "PDB":
-            x = float(row["x"])
-        else:  # mmCIF
-            x = float(row["Cartn_x"])
-        line = line[:30] + f"{x:8.3f}" + line[38:]
+        # --- Format and Write ATOM/HETATM Line ---
+        pdb_line = _format_pdb_atom_line(atom_data)
+        buffer.write(pdb_line + "\n")
 
-        # Set Y coordinate
-        if format_type == "PDB":
-            y = float(row["y"])
-        else:  # mmCIF
-            y = float(row["Cartn_y"])
-        line = line[:38] + f"{y:8.3f}" + line[46:]
-
-        # Set Z coordinate
-        if format_type == "PDB":
-            z = float(row["z"])
-        else:  # mmCIF
-            z = float(row["Cartn_z"])
-        line = line[:46] + f"{z:8.3f}" + line[54:]
-
-        # Set occupancy
-        if format_type == "PDB":
-            occupancy = float(row["occupancy"])
-        else:  # mmCIF
-            occupancy = float(row.get("occupancy", 1.0))
-        line = line[:54] + f"{occupancy:6.2f}" + line[60:]
+        # --- Update Tracking Variables ---
+        last_serial = atom_data["serial"]
+        last_chain_id = current_chain_id
+        last_res_info = current_res_info
 
-        # Set temperature factor
-        if format_type == "PDB":
-            temp_factor = float(row["tempFactor"])
-        else:  # mmCIF
-            temp_factor = float(row.get("B_iso_or_equiv", 0.0))
-        line = line[:60] + f"{temp_factor:6.2f}" + line[66:]
-
-        # Set element symbol
-        if format_type == "PDB":
-            element = row["element"]
-        else:  # mmCIF
-            element = row.get("type_symbol", "")
-        line = line[:76] + element.rjust(2) + line[78:]
-
-        # Set charge
-        if format_type == "PDB":
-            charge = row["charge"]
-        else:  # mmCIF
-            charge = row.get("pdbx_formal_charge", "")
-            if charge and charge not in ["?", "."]:
-                # Convert numeric charge to PDB format (e.g., "1+" or "2-")
-                try:
-                    charge_val = int(charge)
-                    if charge_val != 0:
-                        charge = f"{abs(charge_val)}{'+' if charge_val > 0 else '-'}"
-                    else:
-                        charge = ""
-                except ValueError:
-                    pass
-        line = line[:78] + charge + line[80:]
-
-        # Write the line to the buffer
-        buffer.write(line.rstrip() + "\n")
-
-        # Update last atom info for potential TER record
-        if format_type == "PDB":
-            last_serial = int(row["serial"])
-            last_res_name = row["resName"]
-            last_chain_id = row["chainID"]
-            last_res_seq = str(int(row["resSeq"]))
-            last_icode = row["iCode"] if pd.notna(row["iCode"]) else ""
-        else:  # mmCIF
-            last_serial = int(row["id"])
-            last_res_name = row.get("auth_comp_id", row.get("label_comp_id", ""))
-            last_chain_id = row.get("auth_asym_id", row.get("label_asym_id", ""))
-            last_res_seq = str(int(row.get("auth_seq_id", row.get("label_seq_id", 0))))
-            last_icode = (
-                row.get("pdbx_PDB_ins_code", "")
-                if pd.notna(row.get("pdbx_PDB_ins_code", ""))
-                else ""
-            )
-
-    # Add TER record for the last chain
+    # --- Final Records ---
+    # Add TER record for the very last chain in the last model
     if last_chain_id is not None:
-        # Format TER record according to PDB specification
         ter_serial = str(last_serial + 1).rjust(5)
-        ter_res_name = last_res_name.strip().ljust(3)  # Strip and left-justify
+        ter_res_name = last_res_info[2].strip().rjust(3)
         ter_chain_id = last_chain_id
-        ter_res_seq = last_res_seq.rjust(4)
-        ter_icode = last_icode if last_icode else ""  # Use last recorded iCode
+        ter_res_seq = str(last_res_info[0]).rjust(4)
+        ter_icode = last_res_info[1] if last_res_info[1] else ""
 
-        # Construct the TER line ensuring correct spacing for all fields
         ter_line = f"TER   {ter_serial}      {ter_res_name} {ter_chain_id}{ter_res_seq}{ter_icode}"
         buffer.write(ter_line.ljust(80) + "\n")
 
-    # Add END record
+    # Add ENDMDL if models were used
+    if last_model_num is not None:
+        buffer.write("ENDMDL\n")
+
     buffer.write("END\n")
 
-    # Get the content as a string
+    # --- Output Handling ---
     content = buffer.getvalue()
     buffer.close()
 
-    # Write to output if provided
     if output is not None:
         if isinstance(output, str):
             with open(output, "w") as f:
@@ -463,9 +986,8 @@ def write_pdb(
         else:
             output.write(content)
         return None
-
-    # Return the content as a string
-    return content
+    else:
+        return content
 
 
 def write_cif(
@@ -490,7 +1012,7 @@ def write_cif(
     format_type = df.attrs.get("format", "PDB")
 
     # Create a new DataContainer
-    data_container = DataContainer("data_structure")
+    data_container = DataContainer("rnapolis")
 
     # Define the attributes for atom_site category
     if format_type == "mmCIF":
@@ -519,7 +1041,7 @@ def write_cif(
             "auth_comp_id",  # resName
             "auth_asym_id",  # chainID
             "auth_atom_id",  # name
-            "pdbx_PDB_model_num",  # (generated)
+            "pdbx_PDB_model_num",  # model
         ]
 
     # Prepare rows for the atom_site category
@@ -527,32 +1049,44 @@ def write_cif(
 
     for _, row in df.iterrows():
         if format_type == "mmCIF":
-            # Use existing mmCIF data
-            row_data = [str(row.get(attr, "?")) for attr in attributes]
+            # Use existing mmCIF data, converting None to '?' universally
+            row_data = []
+            for attr in attributes:
+                value = row.get(attr)
+                if pd.isna(value):
+                    # Use '?' as the standard placeholder for missing values
+                    row_data.append("?")
+                else:
+                    # Ensure all non-missing values are converted to string
+                    row_data.append(str(value))
         else:  # PDB format
-            # Map PDB data to mmCIF format
+            # Map PDB data to mmCIF format, converting None to '.' or '?'
             entity_id = "1"  # Default entity ID
-            model_num = "1"  # Default model number
+            model_num = str(int(row["model"]))
+
+            # Pre-process optional fields for mmCIF placeholders
+            element_val = "?" if pd.isna(row.get("element")) else str(row["element"])
+            altloc_val = "." if pd.isna(row.get("altLoc")) else str(row["altLoc"])
+            icode_val = "." if pd.isna(row.get("iCode")) else str(row["iCode"])
+            charge_val = "." if pd.isna(row.get("charge")) else str(row["charge"])
 
             row_data = [
                 str(row["record_type"]),  # group_PDB
                 str(int(row["serial"])),  # id
-                str(row["element"]),  # type_symbol
+                element_val,  # type_symbol
                 str(row["name"]),  # label_atom_id
-                str(row.get("altLoc", "")),  # label_alt_id
+                altloc_val,  # label_alt_id
                 str(row["resName"]),  # label_comp_id
                 str(row["chainID"]),  # label_asym_id
                 entity_id,  # label_entity_id
                 str(int(row["resSeq"])),  # label_seq_id
-                str(row["iCode"])
-                if pd.notna(row["iCode"])
-                else "?",  # pdbx_PDB_ins_code
+                icode_val,  # pdbx_PDB_ins_code
                 f"{float(row['x']):.3f}",  # Cartn_x
                 f"{float(row['y']):.3f}",  # Cartn_y
                 f"{float(row['z']):.3f}",  # Cartn_z
                 f"{float(row['occupancy']):.2f}",  # occupancy
                 f"{float(row['tempFactor']):.2f}",  # B_iso_or_equiv
-                str(row.get("charge", "")) or "?",  # pdbx_formal_charge
+                charge_val,  # pdbx_formal_charge
                 str(int(row["resSeq"])),  # auth_seq_id
                 str(row["resName"]),  # auth_comp_id
                 str(row["chainID"]),  # auth_asym_id

rnapolis/splitter.py

@@ -3,10 +3,14 @@ import argparse
 import os
 import sys
 
-import pandas as pd
-
 from rnapolis.parser import is_cif
-from rnapolis.parser_v2 import parse_cif_atoms, parse_pdb_atoms, write_cif, write_pdb
+from rnapolis.parser_v2 import (
+    fit_to_pdb,
+    parse_cif_atoms,
+    parse_pdb_atoms,
+    write_cif,
+    write_pdb,
+)
 
 
 def main():
@@ -97,12 +101,21 @@ def main():
 
         try:
             if output_format == "PDB":
-                write_pdb(model_df, output_path)
+                df_to_write = fit_to_pdb(model_df)
+                write_pdb(df_to_write, output_path)
             else:  # mmCIF
                 write_cif(model_df, output_path)
+        except ValueError as e:
+            # Handle errors specifically from fit_to_pdb
+            print(
+                f"Error fitting model {model_num} from {args.file} to PDB: {e}. Skipping model.",
+                file=sys.stderr,
+            )
+            continue
         except Exception as e:
+            # Handle general writing errors
             print(
-                f"Error writing file {output_path}: {e}",
+                f"Error writing file {output_path} for model {model_num}: {e}",
                 file=sys.stderr,
             )
             # Optionally continue to next model or exit

rnapolis/unifier.py

@@ -7,7 +7,13 @@ from collections import Counter
 import pandas as pd
 
 from rnapolis.parser import is_cif
-from rnapolis.parser_v2 import parse_cif_atoms, parse_pdb_atoms, write_cif, write_pdb
+from rnapolis.parser_v2 import (
+    fit_to_pdb,
+    parse_cif_atoms,
+    parse_pdb_atoms,
+    write_cif,
+    write_pdb,
+)
 from rnapolis.tertiary_v2 import Structure
 
 
@@ -140,13 +146,22 @@ def main():
 
         ext = ".pdb" if format == "PDB" else ".cif"
 
-        with open(f"{args.output}/{base}{ext}", "w") as f:
-            df = pd.concat([residue.atoms for residue in residues])
+        df = pd.concat([residue.atoms for residue in residues])
 
+        try:
             if format == "PDB":
-                write_pdb(df, f)
+                df_to_write = fit_to_pdb(df)
+                with open(f"{args.output}/{base}{ext}", "w") as f:
+                    write_pdb(df_to_write, f)
             else:
-                write_cif(df, f)
+                with open(f"{args.output}/{base}{ext}", "w") as f:
+                    write_cif(df, f)
+        except ValueError as e:
+            print(
+                f"Error processing {path} for PDB output: {e}. Skipping file.",
+                file=sys.stderr,
+            )
+            continue
 
 
 if __name__ == "__main__":

{rnapolis-0.8.0.dist-info → rnapolis-0.8.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: RNApolis
-Version: 0.8.0
+Version: 0.8.1
 Summary: A Python library containing RNA-related bioinformatics functions and classes
 Home-page: https://github.com/tzok/rnapolis-py
 Author: Tomasz Zok

{rnapolis-0.8.0.dist-info → rnapolis-0.8.1.dist-info}/RECORD

@@ -12,17 +12,17 @@ rnapolis/mmcif_pdbx_v50.dic,sha256=5QFx1ssDaehR4_DQ-tS9VQux262SiLXaqcwmwwejF5c,5
 rnapolis/molecule_filter.py,sha256=jgcpJxx_oXEBX0d30v4k_FdwRouRUPUsEtCYWgLGpD4,7310
 rnapolis/motif_extractor.py,sha256=Lfn1iEkhkP9eZD3GPEWNAfy00QO7QPCc8wM_XS1ory8,1147
 rnapolis/parser.py,sha256=3g4mtFvpiEENFcSBBtx_E_x1vJPF9BujWnts0kb9XjE,16340
-rnapolis/parser_v2.py,sha256=eUccbTXCD5I7q0GVbaGWmjj0CT5d2VK8x9tr0gtrRuA,19801
+rnapolis/parser_v2.py,sha256=qG6CO3or7zmuJu368g9Nzokiqdeip4yjD14F163uH6w,40618
 rnapolis/rfam_folder.py,sha256=SjiiyML_T1__saruFwSMJEoQ7Y55GIU8ktS8ZUn5-fw,11111
-rnapolis/splitter.py,sha256=8mMZ2ZmhqptPUjmkDOFbLvC-dvWpuvJ0beSoeaD5pzk,3642
+rnapolis/splitter.py,sha256=x-Zn21mkiMgvYPptUFD9BbdNIvoaM6b8GzGf6uYXEwE,4052
 rnapolis/tertiary.py,sha256=6t9ZB4w33-5n_M3sns1RoFXCOTgVAgGH4WDNG5OG9Kg,23426
 rnapolis/tertiary_v2.py,sha256=I1uyHWIUePNGO5m-suoL4ibtz02qAJUMvYm0BUKUygY,22480
 rnapolis/transformer.py,sha256=aC0nBmHHJf5TyLvBIV57Jj3tlwpvHbPo347opfAOlQA,3844
-rnapolis/unifier.py,sha256=DR1_IllgaAYT9_FUE6XC9B-2wgqbBHs2D1MjyZT2j2g,5438
+rnapolis/unifier.py,sha256=2ge7IB9FdRgzSAiVD39U_ciwtdDJ2fGzf8mUIudbrqY,5820
 rnapolis/util.py,sha256=IdquFO3PV1_KDqodjupzm0Rqvgy0CeSzxGHaGEHYXVU,543
-rnapolis-0.8.0.dist-info/licenses/LICENSE,sha256=ZGRu12MzCgbYA-Lt8MyBlmjvPZh7xfiD5u5wBx0enq4,1066
-rnapolis-0.8.0.dist-info/METADATA,sha256=zD_byFTP6xNdYCQdu5bslqSE_noBjSagzhn2EOSlcYE,54537
-rnapolis-0.8.0.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
-rnapolis-0.8.0.dist-info/entry_points.txt,sha256=H00KoN54wU3dFOofAu3H_3PADmZOBTB1hXf5TUU2uzo,438
-rnapolis-0.8.0.dist-info/top_level.txt,sha256=LcO18koxZcWoJ21KDRRRo_tyIbmXL5z61dPitZpy8yc,9
-rnapolis-0.8.0.dist-info/RECORD,,
+rnapolis-0.8.1.dist-info/licenses/LICENSE,sha256=ZGRu12MzCgbYA-Lt8MyBlmjvPZh7xfiD5u5wBx0enq4,1066
+rnapolis-0.8.1.dist-info/METADATA,sha256=NOg9-s2n313HElku8z06JiBvEhPf6oV9RR7ur20hwys,54537
+rnapolis-0.8.1.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
+rnapolis-0.8.1.dist-info/entry_points.txt,sha256=H00KoN54wU3dFOofAu3H_3PADmZOBTB1hXf5TUU2uzo,438
+rnapolis-0.8.1.dist-info/top_level.txt,sha256=LcO18koxZcWoJ21KDRRRo_tyIbmXL5z61dPitZpy8yc,9
+rnapolis-0.8.1.dist-info/RECORD,,