def symlink(
source_path,
target_path,
):
rel_symlink(
str(source_path),
str(target_path),
)
def __call__(self, dataset, output_path, regex):
try:
source_ligand_path = self.get_ligand_path(
dataset,
regex,
)
# print(source_ligand_path)
rel_symlink(
str(source_ligand_path),
str(output_path / source_ligand_path.name),
)
except Exception as e:
print(e)
def __call__(self, mcd):
# ==============================>
# Clean previous PanDDA
# ==============================>
try:
shutil.rmtree(self.out_dir, ignore_errors=True)
except Exception as e:
print(e)
# ==============================>
# Make output dir
# ==============================>
os.mkdir(str(self.out_dir))
# ==============================>
# Get path objects
# ==============================>
dataset_template = {"event_map.ccp4": None, "ligand": {"dummy": None}}
processed_datasets = {
dtag: dataset_template
for dtag, d in mcd.datasets.items()
}
analyses = {"pandda_analyse_events.csv": None}
pandda = {
"processed_datasets": processed_datasets,
"analyses": analyses
}
# ==============================>
# Initialise pandda output
# ==============================>
self.tree = Tree(str(self.out_dir), pandda)
# ==============================>
# Initialise dataset output
# ==============================>
errors = []
for dtag, dataset in mcd.datasets.items():
# ==============================>
# Get path
# ==============================>
dataset_path = p.Path(self.tree(("processed_datasets", dtag))[0])
# ==============================>
# Create links to input files
# ==============================>
# Links for the dataset input files
# TODO: seems inelegant
link_pdb = str(dataset_path / "pandda_input.pdb")
link_mtz = str(dataset_path / "pandda_input.mtz")
# Link the input files to the output folder
if not os.path.exists(link_pdb):
rel_symlink(orig=dataset.model.filename, link=link_pdb)
if not os.path.exists(link_mtz):
rel_symlink(orig=dataset.data.filename, link=link_mtz)
# ==============================>
# Search for ligand files and copy them to the output ligands folder
# ==============================>
lig_files = glob.glob(
os.path.join(os.path.dirname(dataset.model.filename),
self.lig_style))
for lig_file in lig_files:
# Find all files with the same basename but allowing for different extensions. Then link to output folder.
lig_base = os.path.splitext(lig_file)[0] + '.*'
lig_matches = glob.glob(lig_base)
for lig in lig_matches:
out_path = os.path.join(str(dataset_path / 'ligand'),
os.path.basename(lig))
if os.path.exists(lig) and (not os.path.exists(out_path)):
try:
shutil.copy(lig, out_path)
except:
pass
# # ==============================>
# # Lastly: Update the pointer to the new path (relative to the pandda directory)
# # ==============================>
# dataset.model.filename = os.path.relpath(link_pdb, start=self.out_dir)
# dataset.data.filename = os.path.relpath(link_mtz, start=self.out_dir)
return self.tree