def test_gds_info(self):
gds_info = GDSInfo()
self.assertIsNotNone(gds_info)
self.assertGreater(len(gds_info.keys()), 0)
self.assertGreater(len(gds_info.items()), 0)
self.assertGreater(len(gds_info.values()), 0)
self.assertIsNotNone(gds_info[self.test_sample])
self.assertEqual(gds_info[self.test_sample]['genes'], 9561)
self.assertEqual(int(gds_info[self.test_sample]['sample_count']), 4)
self.assertEqual(len(gds_info[self.test_sample]['subsets']), 2)
def test_gds_data_transposed(self):
gds_info = GDSInfo()
gds_table = GDS(self.test_sample, transpose=True)
# test if data is downloaded
self.assertIsNotNone(gds_table)
self.assertIsInstance(gds_table, Table)
# test data table values
rows, columns = gds_table.X.shape
self.assertEqual(int(gds_info[self.test_sample]['sample_count']), columns)
self.assertEqual(int(gds_info[self.test_sample]['genes']), rows)
# test data table annotations
self.assertTrue(TableAnnotation.gene_as_attr_name in gds_table.attributes)
self.assertTrue(TableAnnotation.gene_id_column in gds_table.attributes)
self.assertTrue(TableAnnotation.tax_id in gds_table.attributes)
self.assertFalse(gds_table.attributes[TableAnnotation.gene_as_attr_name])
""" Documentation script """
import textwrap
from orangecontrib.bioinformatics.geo.dataset import GDSInfo
gds_info = GDSInfo()
gds = gds_info["GDS10"]
print("ID:")
print(gds["dataset_id"])
print("Features: ")
print(gds["feature_count"])
print("Genes:")
print(gds["gene_count"])
print("Organism:")
print(gds["platform_organism"])
print("PubMed ID:")
print(gds["pubmed_id"])
print("Sample types:")
for sample_type in set([sinfo["type"] for sinfo in gds["subsets"]]):
ss = [
sinfo["description"] for sinfo in gds["subsets"]
if sinfo["type"] == sample_type
]
print(" %s (%s)" % (sample_type, ", ".join(ss)))
print("")
print("Description:")
print("\n".join(textwrap.wrap(gds["description"], 70)))
def get_gds_model(progress=lambda val: None):
"""
Initialize and return a GDS datasets model.
:param progress: A progress callback.
:rval tuple:
A tuple of (QStandardItemModel, GDSInfo, [GDS])
.. note::
The returned QStandardItemModel's thread affinity is set to
the GUI thread.
"""
progress(1)
info = GDSInfo()
search_keys = ["dataset_id", "title", "platform_organism", "description"]
cache_dir = serverfiles.localpath(DOMAIN)
gds_link = "http://www.ncbi.nlm.nih.gov/sites/GDSbrowser?acc={0}"
pm_link = "http://www.ncbi.nlm.nih.gov/pubmed/{0}"
gds_list = []
def is_cached(gds):
return os.path.exists(
os.path.join(cache_dir, gds["dataset_id"]) + ".soft.gz")
def item(displayvalue, item_values={}):
item = QStandardItem()
item.setData(displayvalue, Qt.DisplayRole)
for role, value in item_values.items():
item.setData(value, role)
return item
def gds_to_row(gds):
#: Text for easier full search.
search_text = " | ".join(
[gds.get(key, "").lower() for key in search_keys])
row = [
item(" " if is_cached(gds) else "", {TextFilterRole: search_text}),
item(gds["dataset_id"],
{gui.LinkRole: gds_link.format(gds["dataset_id"])}),
item(gds["title"]),
item(gds["platform_organism"]),
item(len(gds["samples"])),
item(gds["feature_count"]),
item(gds["gene_count"]),
item(len(gds["subsets"])),
item(
gds.get("pubmed_id", ""), {
gui.LinkRole:
pm_link.format(gds["pubmed_id"])
if gds.get("pubmed_id") else None
})
]
return row
model = QStandardItemModel()
model.setHorizontalHeaderLabels([
"", "ID", "Title", "Organism", "Samples", "Features", "Genes",
"Subsets", "PubMedID"
])
progress(20)
for gds in info.values():
model.appendRow(gds_to_row(gds))
gds_list.append(gds)
progress(50)
if QThread.currentThread() is not QCoreApplication.instance().thread():
model.moveToThread(QCoreApplication.instance().thread())
return model, info, gds_list
def valid(info, n=40):
"""Return a set of subset types containing more than n samples in every subset"""
invalid = set()
subsets = set([sinfo["type"] for sinfo in info["subsets"]])
for sampleinfo in info["subsets"]:
if len(sampleinfo["sample_id"]) < n:
invalid.add(sampleinfo["type"])
return subsets.difference(invalid)
def report(stypes, info):
"""Pretty-print GDS and valid susbset types"""
for id, sts in stypes:
print(id)
for st in sts:
gds = info[id]
print(" %s:" % st + ", ".join([
"%s/%d" % (sinfo["description"], len(sinfo["sample_id"]))
for sinfo in gds["subsets"] if sinfo["type"] == st
]))
gdsinfo = GDSInfo()
valid_subset_types = [(id, valid(info)) for id, info in sorted(gdsinfo.items())
if valid(info)]
report(valid_subset_types, gdsinfo)
print('datasets = ' + str(len(valid_subset_types)))
print('type subsets = ' + str(sum(len(b) for _, b in valid_subset_types)))
def __init__(self):
OWWidget.__init__(self)
ConcurrentWidgetMixin.__init__(self)
self.gds_info: Optional[GDSInfo] = GDSInfo(
) # TODO: handle possible exceptions
self.gds_data: Optional[Table] = None
# Control area
box = widgetBox(self.controlArea, 'Info', addSpace=True)
self.infoBox = widgetLabel(box, 'Initializing\n\n')
box = widgetBox(self.controlArea, 'Output', addSpace=True)
radioButtonsInBox(box,
self,
'genes_as_rows',
['Samples in rows', 'Genes in rows'],
callback=self._run)
separator(box)
rubber(self.controlArea)
auto_commit(self.controlArea,
self,
'auto_commit',
'&Commit',
box=False)
# Main Area
# Filter widget
self.filter = lineEdit(self.mainArea,
self,
'search_pattern',
'Filter:',
callbackOnType=True,
callback=self._apply_filter)
self.mainArea.layout().addWidget(self.filter)
splitter_vertical = QSplitter(Qt.Vertical, self.mainArea)
self.mainArea.layout().addWidget(splitter_vertical)
# set table view
self.table_view = QTableView(splitter_vertical)
self.table_view.setShowGrid(False)
self.table_view.setSortingEnabled(True)
self.table_view.sortByColumn(1, Qt.AscendingOrder)
self.table_view.setAlternatingRowColors(True)
self.table_view.verticalHeader().setVisible(False)
self.table_view.setSelectionBehavior(QAbstractItemView.SelectRows)
self.table_view.setSelectionMode(QAbstractItemView.SingleSelection)
self.table_view.viewport().setMouseTracking(True)
self.table_view.setSizeAdjustPolicy(
QAbstractScrollArea.AdjustToContents)
self.table_model = GEODatasetsModel()
self.table_model.initialize(self.gds_info)
self.table_view.setModel(self.table_model)
self.table_model.show_table()
self.table_view.horizontalHeader().setStretchLastSection(True)
self.table_view.resizeColumnsToContents()
v_header = self.table_view.verticalHeader()
option = self.table_view.viewOptions()
size = self.table_view.style().sizeFromContents(
QStyle.CT_ItemViewItem, option, QSize(20, 20), self.table_view)
v_header.setDefaultSectionSize(size.height() + 2)
v_header.setMinimumSectionSize(5)
# set item delegates
self.table_view.setItemDelegateForColumn(
self.table_model.pubmedid_col,
LinkStyledItemDelegate(self.table_view))
self.table_view.setItemDelegateForColumn(
self.table_model.gds_id_col,
LinkStyledItemDelegate(self.table_view))
self.table_view.setItemDelegateForColumn(
self.table_model.indicator_col,
IndicatorItemDelegate(self.table_view, role=Qt.DisplayRole),
)
splitter_horizontal = QSplitter(Qt.Horizontal, splitter_vertical)
# Description Widget
box = widgetBox(splitter_horizontal, 'Description')
self.description_widget = widgetLabel(box, '')
self.description_widget.setWordWrap(True)
rubber(box)
# Sample Annotations Widget
box = widgetBox(splitter_horizontal, 'Sample Annotations')
self.annotations_widget = QTreeWidget(box)
self.annotations_widget.setHeaderLabels(
['Type (Sample annotations)', 'Sample count'])
self.annotations_widget.setRootIsDecorated(True)
box.layout().addWidget(self.annotations_widget)
self._annotations_updating = False
self.annotations_widget.itemChanged.connect(
self.on_annotation_selection_changed)
self.splitters = splitter_vertical, splitter_horizontal
for sp, setting in zip(self.splitters, self.splitter_settings):
sp.splitterMoved.connect(self._splitter_moved)
sp.restoreState(setting)
self.table_view.selectionModel().selectionChanged.connect(
self.on_gds_selection_changed)
self._apply_filter()
self._run()
def setUp(self):
self.info = GDSInfo()