diff --git a/projects/2024-12-ma-denv/example_index.csv.gz b/projects/2024-12-ma-denv/example_index.csv.gz
new file mode 100755
index 0000000000000000000000000000000000000000..23d5092b7f3babf8a77f1c6d7fa886c7adaed01c
Binary files /dev/null and b/projects/2024-12-ma-denv/example_index.csv.gz differ
diff --git a/projects/2024-12-ma-denv/example_result.cif.gz b/projects/2024-12-ma-denv/example_result.cif.gz
new file mode 100644
index 0000000000000000000000000000000000000000..1d078534efb68ebe6516ac4f3f61fe0e7a050c6a
Binary files /dev/null and b/projects/2024-12-ma-denv/example_result.cif.gz differ
diff --git a/projects/2024-12-ma-denv/example_swissmodelcif.cif.gz b/projects/2024-12-ma-denv/example_swissmodelcif.cif.gz
new file mode 100755
index 0000000000000000000000000000000000000000..f76354699f4fe37b517a5ffe546a363b1071f9de
Binary files /dev/null and b/projects/2024-12-ma-denv/example_swissmodelcif.cif.gz differ
diff --git a/projects/2024-12-ma-denv/example_unp_data.json.gz b/projects/2024-12-ma-denv/example_unp_data.json.gz
new file mode 100644
index 0000000000000000000000000000000000000000..35af190837c95979968b3e726939782ce26d4d9b
Binary files /dev/null and b/projects/2024-12-ma-denv/example_unp_data.json.gz differ
diff --git a/projects/2024-12-ma-denv/macif.py b/projects/2024-12-ma-denv/macif.py
new file mode 100644
index 0000000000000000000000000000000000000000..da5b9c46cd2e8a773d717bb46fc38a23c6571e14
--- /dev/null
+++ b/projects/2024-12-ma-denv/macif.py
@@ -0,0 +1,186 @@
+"""mmCIF editing for ModelArchive depositions.
+
+Supposed to look like gemmi.cif but with some convenience on top. Not meant to
+show high-performance but help preparing ModelCIF files for a MA deposition.
+Use for 'one time' jobs, not as frequently run tool in a service/ pipeline."""
+
+import gzip
+
+# For whatever reason, 'no-name-in-module' can not be silenced by config atm
+# pylint: disable=no-name-in-module
+from gemmi import cif
+
+# pylint: enable=no-name-in-module
+import gemmi
+
+
+def _gemmi_quote(value):
+ """Quote string values when necessary."""
+ if (
+ isinstance(value, str)
+ and " " in value
+ and not (value.startswith("'") and value.endswith("'"))
+ and not (value.startswith('"') and value.endswith('"'))
+ ):
+ return cif.quote(value)
+ return value
+
+
+class MABlock:
+ """gemmi.cif wrapper that skips reading/ documents and jumps right into
+ gemmi.cif.Block objects. You have all the gemmi.cif.Block functionality
+ available plus our own convenience functions on top."""
+
+ def __init__(self, model_data):
+ """Initialise a single Block from a file."""
+ self.source = model_data
+ self.doc = cif.read(model_data)
+ self.block = self.doc.sole_block()
+
+ self._targets = None
+ self._polymer_targets = None
+
+ def __getattr__(self, name):
+ """If an attribute is not found, try self.block before exception."""
+ # The catch here is: when asking for self.foo,
+ # self.__getattribute__(self, "foo") is executed first. If "foo" is
+ # not found in self, self.__getattr__(self, "foo") is called. So here
+ # we already know, that "foo" is not there and we can check the
+ # original block.
+ if hasattr(self.block, name):
+ return getattr(self.block, name)
+ raise AttributeError(
+ f"'{type(self).__name__}' object has no attribute '{name}'"
+ )
+
+ @property
+ def targets(self):
+ """Info about targets."""
+ if self._targets is not None:
+ return self._targets
+ self._targets = {}
+ table = self.find("_ma_target_entity.", ["entity_id"])
+ for row in table:
+ if row["entity_id"] in self._targets:
+ raise RuntimeError(
+ f"Target with entity_id '{row['entity_id']}' is duplicated."
+ )
+ self._targets[row["entity_id"]] = {
+ "entity_id": row["entity_id"], # makes live easier for singles
+ "sequence": self.get_sequence(row["entity_id"]),
+ }
+ table = self.find("_entity.", ["id", "type"])
+ for row in table:
+ self._targets[row["id"]]["type"] = row["type"]
+
+ return self._targets
+
+ @property
+ def polymer_targets(self):
+ """Only targets of entity type 'polymer'."""
+ if self._polymer_targets is not None:
+ return self._polymer_targets
+ self._polymer_targets = []
+ for target in self.targets.values():
+ if target["type"] == "polymer":
+ self._polymer_targets.append(target)
+
+ return self._polymer_targets
+
+ def find(self, name, columns):
+ """Get a table with defined colums. Throws an exception if table is not
+ found."""
+ table = self.block.find(name, columns)
+ if len(table) == 0:
+ raise RuntimeError(
+ f"""Table '{name}' with columns '{"', '".join(columns)}' """
+ + "not found."
+ )
+
+ return table
+
+ def get_sequence(self, entity):
+ """Get the sequence of a 'polymer' entity. `entity` is the numeric ID
+ of the entity.
+ Reading sequences is inefficient atm, it reads the whole table for
+ every sequence w/o any caching."""
+ table = self.find("_entity_poly_seq.", ["entity_id", "num", "mon_id"])
+ sequence = ""
+ num = 0
+ for row in table:
+ if row["entity_id"] != entity:
+ continue
+ num += 1
+ assert int(row["num"]) == num
+ sequence += gemmi.find_tabulated_residue(
+ row["mon_id"]
+ ).one_letter_code
+
+ return sequence
+
+ def write_file(self, filename, compress=False, style=cif.Style.Simple):
+ """Write ModelCIF file to disk, compress upon request.
+ Will compress anyways if file ends with '.gz'."""
+ if compress or filename.endswith(".gz"):
+ if not filename.endswith(".gz"):
+ filename += ".gz"
+ with gzip.open(filename, mode="wt", compresslevel=9) as gfh:
+ gfh.write(self.doc.as_string(style))
+ else:
+ self.doc.write_file(filename, style)
+
+ def add_to_category(self, category, match=None, **kwargs):
+ """Add item values to a category.
+ Keyword arguments are reserved for item names."""
+ if category[0] != "_":
+ category = "_" + category
+ if category[-1] != ".":
+ category += "."
+ items = list(kwargs.keys())
+ row = None
+ if match is not None:
+ table = self.find(category, items + [match[0]])
+ for row in table:
+ if row[match[0]] == match[1]:
+ break
+ if row is None:
+ raise RuntimeError(
+ f"No item {match[0]}=={match[1]} found in category "
+ + f"{category}."
+ )
+ else:
+ table = self.find(category, items)
+ assert len(table) == 1
+ row = table[0]
+ for itm, val in kwargs.items():
+ if row[itm] not in [".", "?"]:
+ print(
+ f" replacing '{cif.as_string(row[itm])}' with "
+ + f"'{val}' ({itm})"
+ )
+ row[itm] = _gemmi_quote(val)
+
+ def add_category(self, category, after=None, **kwargs):
+ """Add a new category to the block with only 1 set of values, 1 row
+ thinking of categories as tables. kwargs are reserved for item/ value
+ pairs. after is a special keyword parameter to locate the new category
+ inside the block."""
+ if not category.startswith("_"):
+ category = "_" + category
+ # handle quoting
+ for values in kwargs.values():
+ for i, val in enumerate(values):
+ values[i] = _gemmi_quote(val)
+ self.block.set_mmcif_category(category, kwargs, raw=True)
+
+ if after is None:
+ return
+ if not after.startswith("_"):
+ after = "_" + after
+ if not after.endswith("."):
+ after += "."
+ table = self.block.find_mmcif_category(after)
+ idx = self.block.get_index(table.tags[-1])
+ # be careful with move_item: loops are 1 item and move as a whole,
+ # single line values move per item/ value par.
+ self.block.move_item(-1, idx + 1)
diff --git a/projects/2024-12-ma-denv/translate2modelcif.py b/projects/2024-12-ma-denv/translate2modelcif.py
index aa3a49baf05b6b4f5bccdde84ea9563b86eebad6..9fca2a7d8c4ee2ee5c5adfc088f3511170d7b0a2 100644
--- a/projects/2024-12-ma-denv/translate2modelcif.py
+++ b/projects/2024-12-ma-denv/translate2modelcif.py
@@ -3,30 +3,26 @@
"""Check & extend SWISS-MODEL models for ma-denv
Example for running:
-ToDo: Example call
-
-ToDo: Update expected output once script is done
-Expected output in ./modelcif for example above:
-- ma-taas-0272.cif as ModelCIF file
-- ma-taas-0272.zip with accompanying data
-- ma-taas-0272-image.png as image to use in ModelArchive
-- ma-taas-0272-issues.json listing issues for that conversion (if any)
+python translate2modelcif.py --unp-json example_unp_data.json \
+ example_swissmodelcif.cif.gz \
+ example_index.csv \
+ done
+
+This call uses the example data from this directory. You may need to decompress
+some of the files. The resulting './done/example_swissmodelcif.cif.gz' should
+be identical to './example_swissmodelcif.cif.gz'.
+
+Expected output in ./done for example above:
+- example_swissmodelcif.cif.gz as ModelCIF file with extended annotation
"""
from datetime import date
import argparse
import csv
-import gzip
import os
import sys
-# For whatever reason, 'no-name-in-module' can not be silenced by config atm
-# pylint: disable=no-name-in-module
-from gemmi import cif
-
-# pylint: enable=no-name-in-module
-import gemmi
-
+from macif import MABlock
from uniprotkb import UniProtKBEntryCache
@@ -84,7 +80,6 @@ def _parse_args():
metavar="<INPUT MODELCIF FILE>",
help="Path to a SWISS-MODEL ModelCIF file provided by depositors.",
)
- # ToDo: fill '?'
parser.add_argument(
"model_info_csv",
type=str,
@@ -183,7 +178,7 @@ def _update_modelcif(mdl_file, mdl_info, unp_json_file, out_dir, compress):
Caution: This is for updates BEFORE deposition, this update does not do
a CIF-style revision history for you!"""
- block = _MABlock(mdl_file)
+ block = MABlock(mdl_file)
assert len(block.polymer_targets) == 1
target = block.polymer_targets[0]
unp = UniProtKBEntryCache(unp_json_file)
@@ -253,178 +248,6 @@ def _update_modelcif(mdl_file, mdl_info, unp_json_file, out_dir, compress):
)
-################################################################################
-# FUNCTIONALITY THAT MAY MOVE INTO ITS OWN MODULE
-################################################################################
-
-
-class _MABlock:
- """gemmi.cif wrapper that skips reading/ documents and jumps right into
- gemmi.cif.Block objects. You have all the gemmi.cif.Block functionality
- available plus our own convenience functions on top."""
-
- def __init__(self, model_data):
- """Initialise a single Block from a file."""
- self.source = model_data
- self.doc = cif.read(model_data)
- self.block = self.doc.sole_block()
-
- self._targets = None
- self._polymer_targets = None
-
- def __getattr__(self, name):
- """If an attribute is not found, try self.block before exception."""
- # The catch here is: when asking for self.foo,
- # self.__getattribute__(self, "foo") is executed first. If "foo" is
- # not found in self, self.__getattr__(self, "foo") is called. So here
- # we already know, that "foo" is not there and we can check the
- # original block.
- if hasattr(self.block, name):
- return getattr(self.block, name)
- raise AttributeError(
- f"'{type(self).__name__}' object has no attribute '{name}'"
- )
-
- @property
- def targets(self):
- """Info about targets."""
- if self._targets is not None:
- return self._targets
- self._targets = {}
- table = self.find("_ma_target_entity.", ["entity_id"])
- for row in table:
- if row["entity_id"] in self._targets:
- raise RuntimeError(
- f"Target with entity_id '{row['entity_id']}' is duplicated."
- )
- self._targets[row["entity_id"]] = {
- "entity_id": row["entity_id"], # makes live easier for singles
- "sequence": self.get_sequence(row["entity_id"]),
- }
- table = self.find("_entity.", ["id", "type"])
- for row in table:
- self._targets[row["id"]]["type"] = row["type"]
-
- return self._targets
-
- @property
- def polymer_targets(self):
- """Only targets of entity type 'polymer'."""
- if self._polymer_targets is not None:
- return self._polymer_targets
- self._polymer_targets = []
- for target in self.targets.values():
- if target["type"] == "polymer":
- self._polymer_targets.append(target)
-
- return self._polymer_targets
-
- def find(self, name, columns):
- """Get a table with defined colums. Throws an exception if table is not
- found."""
- table = self.block.find(name, columns)
- if len(table) == 0:
- raise RuntimeError(
- f"""Table '{name}' with columns '{"', '".join(columns)}' """
- + "not found."
- )
-
- return table
-
- def get_sequence(self, entity):
- """Get the sequence of a 'polymer' entity. `entity` is the numeric ID
- of the entity.
- Reading sequences is inefficient atm, it reads the whole table for
- every sequence w/o any caching."""
- table = self.find("_entity_poly_seq.", ["entity_id", "num", "mon_id"])
- sequence = ""
- num = 0
- for row in table:
- if row["entity_id"] != entity:
- continue
- num += 1
- assert int(row["num"]) == num
- sequence += gemmi.find_tabulated_residue(
- row["mon_id"]
- ).one_letter_code
-
- return sequence
-
- def write_file(self, filename, compress=False, style=cif.Style.Simple):
- """Write ModelCIF file to disk, compress upon request.
- Will compress anyways if file ends with '.gz'."""
- if compress or filename.endswith(".gz"):
- if not filename.endswith(".gz"):
- filename += ".gz"
- with gzip.open(filename, mode="wt", compresslevel=9) as gfh:
- gfh.write(self.doc.as_string(style))
- else:
- self.doc.write_file(filename, style)
-
- def add_to_category(self, category, match=None, **kwargs):
- """Add item values to a category.
- Keyword arguments are reserved for item names."""
- if category[0] != "_":
- category = "_" + category
- if category[-1] != ".":
- category += "."
- items = list(kwargs.keys())
- row = None
- if match is not None:
- table = self.find(category, items + [match[0]])
- for row in table:
- if row[match[0]] == match[1]:
- break
- if row is None:
- raise RuntimeError(
- f"No item {match[0]}=={match[1]} found in category "
- + f"{category}."
- )
- else:
- table = self.find(category, items)
- assert len(table) == 1
- row = table[0]
- for itm, val in kwargs.items():
- if row[itm] not in [".", "?"]:
- print(
- f" replacing '{cif.as_string(row[itm])}' with "
- + f"'{val}' ({itm})"
- )
- # ToDo: handle quoting the same way as for add_category
- row[itm] = cif.quote(val) if " " in val else val
-
- def add_category(self, category, after=None, **kwargs):
- """Add a new category to the block with only 1 set of values, 1 row
- thinking of categories as tables. kwargs are reserved for item/ value
- pairs. after is a special keyword parameter to locate the new category
- inside the block."""
- if not category.startswith("_"):
- category = "_" + category
- # handle quoting
- for values in kwargs.values():
- for i, val in enumerate(values):
- if (
- isinstance(val, str)
- and " " in val
- and not (val.startswith("'") and val.endswith("'"))
- and not (val.startswith('"') and val.endswith('"'))
- ):
- values[i] = cif.quote(val)
- self.block.set_mmcif_category(category, kwargs, raw=True)
-
- if after is None:
- return
- if not after.startswith("_"):
- after = "_" + after
- if not after.endswith("."):
- after += "."
- table = self.block.find_mmcif_category(after)
- idx = self.block.get_index(table.tags[-1])
- # be careful with move_item: loops are 1 item and move as a whole,
- # single line values move per item/ value par.
- self.block.move_item(-1, idx + 1)
-
-
################################################################################