def send_to_output(self, result):
self.progress_bar.finish()
self.setStatusMessage('')
etc_json, table_name = result
# convert to table
data = etc_to_table(etc_json, bool(self.gene_as_attr_name))
# set table name
data.name = table_name
# match genes
gene_matcher = GeneMatcher(str(self.organism))
if not bool(self.gene_as_attr_name):
if 'Gene' in data.domain:
data = gene_matcher.match_table_column(
data, 'Gene', StringVariable(ENTREZ_ID))
data.attributes[GENE_ID_COLUMN] = ENTREZ_ID
else:
gene_matcher.match_table_attributes(data)
data.attributes[GENE_ID_ATTRIBUTE] = ENTREZ_ID
# add table attributes
data.attributes[TAX_ID] = str(self.organism)
data.attributes[GENE_AS_ATTRIBUTE_NAME] = bool(self.gene_as_attr_name)
# reset cache indicators
self.set_cached_indicator()
# send data to the output signal
self.Outputs.etc_data.send(data)
def test_match_table_attributes(self):
gm = GeneMatcher('4932')
data = Table('brown-selected.tab')
data = Table.transpose(data, feature_names_column='gene')
gm.match_table_attributes(data)
for column in data.domain.attributes:
self.assertTrue(ENTREZ_ID in column.attributes)
def send_to_output(self, result):
self.progress_bar.finish()
self.setStatusMessage('')
etc_json, table_name = result
# convert to table
data = etc_to_table(etc_json, bool(self.gene_as_attr_name))
# set table name
data.name = table_name
# match genes
gene_matcher = GeneMatcher(str(self.orgnism))
if not bool(self.gene_as_attr_name):
if 'Gene' in data.domain:
gene_column = data.domain['Gene']
gene_names = data.get_column_view(gene_column)[0]
gene_matcher.genes = gene_names
gene_matcher.run_matcher()
domain_ids = Domain([], metas=[StringVariable(NCBI_ID)])
data_ids = [[str(gene.ncbi_id) if gene.ncbi_id else '?']
for gene in gene_matcher.genes]
table_ids = Table(domain_ids, data_ids)
data = Table.concatenate([data, table_ids])
data.attributes[GENE_ID_COLUMN] = NCBI_ID
else:
gene_matcher.match_table_attributes(data)
data.attributes[GENE_ID_ATTRIBUTE] = NCBI_ID
# add table attributes
data.attributes[TAX_ID] = str(self.orgnism)
data.attributes[GENE_AS_ATTRIBUTE_NAME] = bool(self.gene_as_attr_name)
# reset cache indicators
self.set_cached_indicator()
# send data to the output signal
self.Outputs.etc_data.send(data)
class OWGenes(OWWidget, ConcurrentWidgetMixin):
name = "Genes"
description = "Tool for working with genes"
icon = "../widgets/icons/OWGeneInfo.svg"
priority = 40
want_main_area = True
selected_organism: int = Setting(11)
search_pattern: str = Setting('')
exclude_unmatched = Setting(True)
replace_id_with_symbol = Setting(True)
auto_commit = Setting(True)
settingsHandler = DomainContextHandler()
selected_gene_col = ContextSetting(None)
use_attr_names = ContextSetting(True)
replaces = ['orangecontrib.bioinformatics.widgets.OWGeneNameMatcher.OWGeneNameMatcher']
class Inputs:
data_table = Input("Data", Table)
class Outputs:
data_table = Output("Data", Table)
gene_matcher_results = Output("Genes", Table)
class Information(OWWidget.Information):
pass
def sizeHint(self):
return QSize(1280, 960)
def __init__(self):
OWWidget.__init__(self)
ConcurrentWidgetMixin.__init__(self)
# ATTRIBUTES #
self.target_database = ENTREZ_ID
# input data
self.input_data = None
self.input_genes = None
self.tax_id = None
self.column_candidates = []
# input options
self.organisms = []
# gene matcher
self.gene_matcher = None
# progress bar
self.progress_bar = None
self._timer = QTimer()
self._timer.timeout.connect(self._apply_filter)
self._timer.setSingleShot(True)
# GUI SECTION #
# Control area
self.info_box = widgetLabel(widgetBox(self.controlArea, "Info", addSpace=True), 'No data on input.\n')
organism_box = vBox(self.controlArea, 'Organism')
self.organism_select_combobox = comboBox(
organism_box, self, 'selected_organism', callback=self.on_input_option_change
)
self.get_available_organisms()
self.organism_select_combobox.setCurrentIndex(self.selected_organism)
box = widgetBox(self.controlArea, 'Gene IDs in the input data')
self.gene_columns_model = itemmodels.DomainModel(valid_types=(StringVariable, DiscreteVariable))
self.gene_column_combobox = comboBox(
box,
self,
'selected_gene_col',
label='Stored in data column',
model=self.gene_columns_model,
sendSelectedValue=True,
callback=self.on_input_option_change,
)
self.attr_names_checkbox = checkBox(
box,
self,
'use_attr_names',
'Stored as feature (column) names',
disables=[(-1, self.gene_column_combobox)],
callback=self.on_input_option_change,
)
self.gene_column_combobox.setDisabled(bool(self.use_attr_names))
output_box = vBox(self.controlArea, 'Output')
# separator(output_box)
# output_box.layout().addWidget(horizontal_line())
# separator(output_box)
self.exclude_radio = checkBox(
output_box, self, 'exclude_unmatched', 'Exclude unmatched genes', callback=self.commit
)
self.replace_radio = checkBox(
output_box, self, 'replace_id_with_symbol', 'Replace feature IDs with gene names', callback=self.commit
)
auto_commit(self.controlArea, self, "auto_commit", "&Commit", box=False)
rubber(self.controlArea)
# Main area
self.filter = lineEdit(
self.mainArea, self, 'search_pattern', 'Filter:', callbackOnType=True, callback=self.handle_filter_callback
)
# rubber(self.radio_group)
self.mainArea.layout().addWidget(self.filter)
# set splitter
self.splitter = QSplitter()
self.splitter.setOrientation(Qt.Vertical)
self.table_model = GeneInfoModel()
self.table_view = QTableView()
self.table_view.setAlternatingRowColors(True)
self.table_view.viewport().setMouseTracking(True)
self.table_view.setSortingEnabled(True)
self.table_view.setShowGrid(False)
self.table_view.verticalHeader().hide()
# self.table_view.horizontalHeader().setSectionResizeMode(QHeaderView.Stretch)
self.unknown_model = UnknownGeneInfoModel()
self.unknown_view = QTableView()
self.unknown_view.setModel(self.unknown_model)
self.unknown_view.verticalHeader().hide()
self.unknown_view.setShowGrid(False)
self.unknown_view.setSelectionMode(QAbstractItemView.NoSelection)
self.unknown_view.horizontalHeader().setSectionResizeMode(QHeaderView.Stretch)
self.splitter.addWidget(self.table_view)
self.splitter.addWidget(self.unknown_view)
self.splitter.setStretchFactor(0, 90)
self.splitter.setStretchFactor(1, 10)
self.mainArea.layout().addWidget(self.splitter)
def handle_filter_callback(self):
self._timer.stop()
self._timer.start(500)
def _apply_filter(self):
# filter only if input data is present and model is populated
if self.table_model.table is not None:
self.table_model.update_model(filter_pattern=str(self.search_pattern))
self.commit()
def __reset_widget_state(self):
self.table_view.clearSpans()
self.table_view.setModel(None)
self.table_model.clear()
self.unknown_model.clear()
self._update_info_box()
def _update_info_box(self):
if self.input_genes and self.gene_matcher:
num_genes = len(self.gene_matcher.genes)
known_genes = len(self.gene_matcher.get_known_genes())
info_text = (
'{} genes in input data\n'
'{} genes match Entrez database\n'
'{} genes with match conflicts\n'.format(num_genes, known_genes, num_genes - known_genes)
)
else:
info_text = 'No data on input.'
self.info_box.setText(info_text)
def on_done(self, _):
# update info box
self._update_info_box()
# set output options
self.toggle_radio_options()
# set known genes
self.table_model.initialize(self.gene_matcher.genes)
self.table_view.setModel(self.table_model)
self.table_view.selectionModel().selectionChanged.connect(self.commit)
self.table_view.setSelectionBehavior(QAbstractItemView.SelectRows)
self.table_view.setItemDelegateForColumn(
self.table_model.entrez_column_index, LinkStyledItemDelegate(self.table_view)
)
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)
self.table_view.horizontalHeader().setStretchLastSection(True)
# set unknown genes
self.unknown_model.initialize(self.gene_matcher.genes)
self.unknown_view.verticalHeader().setStretchLastSection(True)
self._apply_filter()
def get_available_organisms(self):
available_organism = sorted(
((tax_id, taxonomy.name(tax_id)) for tax_id in taxonomy.common_taxids()), key=lambda x: x[1]
)
self.organisms = [tax_id[0] for tax_id in available_organism]
self.organism_select_combobox.addItems([tax_id[1] for tax_id in available_organism])
def gene_names_from_table(self):
"""Extract and return gene names from `Orange.data.Table`."""
self.input_genes = []
if self.input_data:
if self.use_attr_names:
self.input_genes = [str(attr.name).strip() for attr in self.input_data.domain.attributes]
else:
if self.selected_gene_col is None:
self.selected_gene_col = self.gene_column_identifier()
self.input_genes = [
str(e[self.selected_gene_col]) for e in self.input_data if not np.isnan(e[self.selected_gene_col])
]
def _update_gene_matcher(self):
self.gene_names_from_table()
self.gene_matcher = GeneMatcher(self.get_selected_organism(), auto_start=False)
self.gene_matcher.genes = self.input_genes
# self.gene_matcher.organism = self.get_selected_organism()
def get_selected_organism(self):
return self.organisms[self.selected_organism]
def _run(self):
if self.gene_matcher is not None:
self.start(run_gene_matcher, self.gene_matcher)
def on_input_option_change(self):
self.__reset_widget_state()
self._update_gene_matcher()
self._run()
def gene_column_identifier(self):
"""
Get most suitable column that stores genes. If there are
several suitable columns, select the one with most unique
values. Take the best one.
"""
# candidates -> (variable, num of unique values)
candidates = (
(col, np.unique(self.input_data.get_column_view(col)[0]).size)
for col in self.gene_columns_model
if isinstance(col, DiscreteVariable) or isinstance(col, StringVariable)
)
best_candidate, _ = sorted(candidates, key=lambda x: x[1])[-1]
return best_candidate
def find_genes_location(self):
"""Try locate the genes in the input data when we first load the data.
Proposed rules:
- when no suitable feature names are present, check the columns.
- find the most suitable column, that is, the one with most unique values.
"""
domain = self.input_data.domain
if not domain.attributes:
if self.selected_gene_col is None:
self.selected_gene_col = self.gene_column_identifier()
self.use_attr_names = False
@Inputs.data_table
def handle_input(self, data):
self.closeContext()
self.input_data = None
self.input_genes = None
self.__reset_widget_state()
self.gene_columns_model.set_domain(None)
self.selected_gene_col = None
if data:
self.input_data = data
self.gene_columns_model.set_domain(self.input_data.domain)
# check if input table has tax_id, human is used if tax_id is not found
self.tax_id = str(self.input_data.attributes.get(TAX_ID, '9606'))
# check for gene location. Default is that genes are attributes in the input table.
self.use_attr_names = self.input_data.attributes.get(GENE_AS_ATTRIBUTE_NAME, self.use_attr_names)
if self.tax_id in self.organisms and not self.selected_organism:
self.selected_organism = self.organisms.index(self.tax_id)
self.openContext(self.input_data.domain)
self.find_genes_location()
self.on_input_option_change()
def commit(self):
selection = self.table_view.selectionModel().selectedRows(self.table_model.entrez_column_index)
selected_genes = [row.data() for row in selection]
if not len(selected_genes):
selected_genes = self.table_model.get_filtered_genes()
gene_ids = self.get_target_ids()
known_genes = [gid for gid in gene_ids if gid != '?']
table = None
gm_table = None
if known_genes:
# Genes are in rows (we have a column with genes).
if not self.use_attr_names:
if self.target_database in self.input_data.domain:
gene_var = self.input_data.domain[self.target_database]
metas = self.input_data.domain.metas
else:
gene_var = StringVariable(self.target_database)
metas = self.input_data.domain.metas + (gene_var,)
domain = Domain(self.input_data.domain.attributes, self.input_data.domain.class_vars, metas)
table = self.input_data.transform(domain)
col, _ = table.get_column_view(gene_var)
col[:] = gene_ids
# filter selected rows
selected_genes_set = set(selected_genes)
selected_rows = [
row_index for row_index, row in enumerate(table) if str(row[gene_var]) in selected_genes_set
]
# handle table attributes
table.attributes[TAX_ID] = self.get_selected_organism()
table.attributes[GENE_AS_ATTRIBUTE_NAME] = False
table.attributes[GENE_ID_COLUMN] = self.target_database
table = table[selected_rows] if selected_rows else table
if self.exclude_unmatched:
# create filter from selected column for genes
only_known = table_filter.FilterStringList(gene_var, known_genes)
# apply filter to the data
table = table_filter.Values([only_known])(table)
self.Outputs.data_table.send(table)
# genes are are in columns (genes are features).
else:
domain = self.input_data.domain.copy()
table = self.input_data.transform(domain)
table = self.gene_matcher.match_table_attributes(table, run=False, rename=self.replace_id_with_symbol)
# filter selected columns
selected_genes_set = set(selected_genes)
selected = [
column
for column in table.domain.attributes
if self.target_database in column.attributes
and str(column.attributes[self.target_database]) in selected_genes_set
]
output_attrs = table.domain.attributes
if selected:
output_attrs = selected
if self.exclude_unmatched:
known_genes_set = set(known_genes)
output_attrs = [
col for col in output_attrs if col.attributes[self.target_database] in known_genes_set
]
domain = Domain(output_attrs, table.domain.class_vars, table.domain.metas)
table = table.from_table(domain, table)
# handle table attributes
table.attributes[TAX_ID] = self.get_selected_organism()
table.attributes[GENE_AS_ATTRIBUTE_NAME] = True
table.attributes[GENE_ID_ATTRIBUTE] = self.target_database
gm_table = self.gene_matcher.to_data_table(selected_genes=selected_genes if selected_genes else None)
self.Outputs.data_table.send(table)
self.Outputs.gene_matcher_results.send(gm_table)
def toggle_radio_options(self):
self.replace_radio.setEnabled(bool(self.use_attr_names))
if self.gene_matcher.genes:
# enable checkbox if unknown genes are detected
self.exclude_radio.setEnabled(len(self.gene_matcher.genes) != len(self.gene_matcher.get_known_genes()))
self.exclude_unmatched = len(self.gene_matcher.genes) != len(self.gene_matcher.get_known_genes())
def get_target_ids(self):
return [str(gene.gene_id) if gene.gene_id else '?' for gene in self.gene_matcher.genes]
def runner(self, state: TaskState) -> Table:
exp_type = self.data_output_options.expression_type[self.exp_type].type
exp_source = self.data_output_options.expression_sources[
self.exp_source]
proc_slug = self.data_output_options.process[self.proc_slug].slug
collection_id = self.selected_collection_id
table = self.data_table
progress_steps_download = iter(np.linspace(0, 50, 2))
def callback(i: float, status=""):
state.set_progress_value(i * 100)
if status:
state.set_status(status)
if state.is_interruption_requested():
raise Exception
if not table:
collection = self.res.get_collection_by_id(collection_id)
coll_table = resdk.tables.RNATables(
collection,
expression_source=exp_source,
expression_process_slug=proc_slug,
progress_callable=wrap_callback(callback, end=0.5),
)
species = coll_table._data[0].output['species']
sample = coll_table._samples[0]
state.set_status('Downloading ...')
loop = asyncio.new_event_loop()
asyncio.set_event_loop(loop)
df_exp = coll_table.exp if exp_type != 'rc' else coll_table.rc
df_exp = df_exp.rename(index=coll_table.readable_index)
df_metas = coll_table.meta
df_metas = df_metas.rename(index=coll_table.readable_index)
df_qc = None
if self.append_qc_data:
# TODO: check if there is a way to detect if collection
# table contains QC data
try:
df_qc = coll_table.qc
df_qc = df_qc.rename(index=coll_table.readable_index)
except ValueError:
pass
loop.close()
state.set_status('To data table ...')
duplicates = {
item
for item, count in Counter([
label.split('.')[1]
for label in df_metas.columns.to_list() if '.' in label
]).items() if count > 1
}
# what happens if there is more nested sections?
section_name_to_label = {
section['name']: section['label']
for section in sample.descriptor_schema.schema
}
column_labels = {}
for field_schema, fields, path in iterate_schema(
sample.descriptor, sample.descriptor_schema.schema,
path=''):
path = path[1:] # this is ugly, but cant go around it
if path not in df_metas.columns:
continue
label = field_schema['label']
section_name, field_name = path.split('.')
column_labels[path] = (
label if field_name not in duplicates else
f'{section_name_to_label[section_name]} - {label}')
df_exp = df_exp.reset_index(drop=True)
df_metas = df_metas.astype('object')
df_metas = df_metas.fillna(np.nan)
df_metas = df_metas.replace('nan', np.nan)
df_metas = df_metas.rename(columns=column_labels)
if df_qc is not None:
df_metas = pd.merge(df_metas,
df_qc,
left_index=True,
right_index=True)
xym, domain_metas = vars_from_df(df_metas)
x, _, m = xym
x_metas = np.hstack((x, m))
attrs = [ContinuousVariable(col) for col in df_exp.columns]
metas = domain_metas.attributes + domain_metas.metas
domain = Domain(attrs, metas=metas)
table = Table(domain, df_exp.to_numpy(), metas=x_metas)
state.set_progress_value(next(progress_steps_download))
state.set_status('Matching genes ...')
progress_steps_gm = iter(
np.linspace(50, 99, len(coll_table.gene_ids)))
def gm_callback():
state.set_progress_value(next(progress_steps_gm))
tax_id = species_name_to_taxid(species)
gm = GeneMatcher(tax_id, progress_callback=gm_callback)
table = gm.match_table_attributes(table, rename=True)
table.attributes[TableAnnotation.tax_id] = tax_id
table.attributes[TableAnnotation.gene_as_attr_name] = True
table.attributes[TableAnnotation.gene_id_attribute] = 'Entrez ID'
self.data_table = table
state.set_status('Normalizing ...')
table = self.normalize(table)
state.set_progress_value(100)
return table
def runner(
res: ResolweAPI,
data_objects: List[Data],
options: DataOutputOptions,
exp_type: int,
proc_type: int,
input_annotation: int,
state: TaskState,
) -> Table:
data_frames = []
metadata = defaultdict(list)
def parse_sample_descriptor(sample: Sample) -> None:
general = sample.descriptor.get('general', {})
for label in SAMPLE_DESCRIPTOR_LABELS:
metadata[label].append([general.get(label, '')])
metadata['sample_name'].append([sample.name])
exp_type = file_output_field = options.expression[exp_type].type
proc_type = options.process[proc_type].type
source = options.input_annotation[input_annotation].source
species = options.input_annotation[input_annotation].species
build = options.input_annotation[input_annotation].build
# apply filters
data_objects = [obj for obj in data_objects if obj.process.type == proc_type]
data_objects = [
obj
for obj in data_objects
if obj.output['source'] == source and obj.output['species'] == species and obj.output['build'] == build
]
if exp_type != 'rc':
file_output_field = 'exp'
data_objects = [obj for obj in data_objects if obj.output['exp_type'] == exp_type]
if not data_objects:
raise ResolweDataObjectsNotFound
step, steps = 0, len(data_objects) + 3
def set_progress():
nonlocal step
step += 1
state.set_progress_value(100 * (step / steps))
state.set_status('Downloading ...')
for data_object in data_objects:
set_progress()
parse_sample_descriptor(data_object.sample)
metadata['expression_type'].append([exp_type.upper()])
response = res.get_expressions(data_object.id, data_object.output[file_output_field]['file'])
with io.BytesIO() as f:
f.write(response.content)
f.seek(0)
# expressions to data frame
df = pd.read_csv(f, sep='\t', compression='gzip')
df = df.set_index('Gene').T.reset_index(drop=True)
data_frames.append(df)
state.set_status('Concatenating samples ...')
df = pd.concat(data_frames, axis=0)
state.set_status('To data table ...')
table = table_from_frame(df)
set_progress()
state.set_status('Adding metadata ...')
metas = [StringVariable(label) for label in metadata.keys()]
domain = Domain(table.domain.attributes, table.domain.class_vars, metas)
table = table.transform(domain)
for key, value in metadata.items():
table[:, key] = value
set_progress()
state.set_status('Matching genes ...')
tax_id = species_name_to_taxid(species)
gm = GeneMatcher(tax_id)
table = gm.match_table_attributes(table, rename=True)
table.attributes[TableAnnotation.tax_id] = tax_id
table.attributes[TableAnnotation.gene_as_attr_name] = True
table.attributes[TableAnnotation.gene_id_attribute] = 'Entrez ID'
set_progress()
return table
def _on_dataready(self):
self.setEnabled(True)
self.setBlocking(False)
self.progressBarFinished(processEvents=False)
try:
data = self._datatask.result()
except urlrequest.URLError as error:
self.error(0, ("Error while connecting to the NCBI ftp server! "
"'%s'" % error))
sys.excepthook(type(error), error, getattr(error, "__traceback__"))
return
finally:
self._datatask = None
data_name = data.name
samples, _ = self.selectedSamples()
self.warning(0)
message = None
from orangecontrib.bioinformatics.ncbi.gene import GeneMatcher
gene_matcher = GeneMatcher(self.currentGds.get('taxid', ''))
if self.outputRows:
def samplesinst(ex):
out = []
for meta in data.domain.metas:
out.append((meta.name, ex[meta].value))
if data.domain.class_var.name != 'class':
out.append((data.domain.class_var.name,
ex[data.domain.class_var].value))
return out
samples = set(samples)
mask = [samples.issuperset(samplesinst(ex)) for ex in data]
data = data[numpy.array(mask, dtype=bool)]
gene_matcher.match_table_attributes(data)
if len(data) == 0:
message = "No samples with selected sample annotations."
else:
samples = set(samples)
domain = Domain(
[attr for attr in data.domain.attributes
if samples.issuperset(attr.attributes.items())],
data.domain.class_var,
data.domain.metas
)
# domain.addmetas(data.domain.getmetas())
if len(domain.attributes) == 0:
message = "No samples with selected sample annotations."
stypes = set(s[0] for s in samples)
for attr in domain.attributes:
attr.attributes = dict(
(key, value) for key, value in attr.attributes.items()
if key in stypes
)
data = Table(domain, data)
if 'gene' in data.domain:
gene_column = data.domain['gene']
gene_names = data.get_column_view(gene_column)[0]
gene_matcher.genes = gene_names
gene_matcher.run_matcher()
domain_ids = Domain([], metas=[StringVariable(NCBI_ID)])
data_ids = [[str(gene.ncbi_id) if gene.ncbi_id else '?'] for gene in gene_matcher.genes]
table_ids = Table(domain_ids, data_ids)
data = Table.concatenate([data, table_ids])
if message is not None:
self.warning(0, message)
data.attributes[TAX_ID] = self.currentGds.get('taxid', '')
data.attributes[GENE_AS_ATTRIBUTE_NAME] = bool(self.outputRows)
if not bool(self.outputRows):
data.attributes[GENE_ID_COLUMN] = NCBI_ID
else:
data.attributes[GENE_ID_ATTRIBUTE] = NCBI_ID
data.name = data_name
self.send("Expression Data", data)
model = self.treeWidget.model().sourceModel()
row = self.gds.index(self.currentGds)
model.setData(model.index(row, 0), " ", Qt.DisplayRole)
self.updateInfo()
self.selectionChanged = False
