def generate_project_similarity_plots(self):
plots = []
identifier = "Similarities"
title = "Similarities to other Projects"
plots.append(
viz.get_table(self.similar_projects, identifier + ' table',
title + ' table'))
plots.append(
viz.get_sankey_plot(self.similar_projects,
identifier,
args={
'source': 'current',
'target': 'other',
'weight': 'similarity_pearson',
'orientation': 'h',
'valueformat': '.0f',
'width': 800,
'height': 800,
'font': 12,
'title': title
}))
plots.append(self.get_similarity_network())
return plots
def get_project_information(driver, project_id):
query_name = 'project_graph'
res = pd.DataFrame()
try:
query = ''
parameters = {'project_id': project_id}
data_upload_cypher = get_data_upload_queries()
queries = data_upload_cypher[query_name]['query'].split(';')[:-1]
for query in queries:
res = connector.getCursorData(driver,
query + ';',
parameters=parameters)
except Exception as err:
exc_type, exc_obj, exc_tb = sys.exc_info()
fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
logger.error(
"Error: {}. Creating analytical samples: Query name ({}) - Query ({}), error info: {}, file: {},line: {}"
.format(err, query_name, query, sys.exc_info(), fname,
exc_tb.tb_lineno))
if not res.empty:
res = viz.get_table(
res,
identifier='new_project',
title='Data Uploaded for Project {}'.format(project_id))
return res
def generate_project_attributes_plot(self):
project_df = self.to_dataframe()
project_df = project_df.drop(['similar_projects', 'overlap'], axis=1)
identifier = "Project info"
title = "Project: {} information".format(self.name)
plot = [viz.get_table(project_df, identifier, title)]
return plot
def get_knowledge_graph_plot(self):
if self.graph is None:
self.generate_knowledge_graph()
title = 'Project {} Knowledge Graph'.format(self.identifier)
if self.data is not None:
if 'name' in self.data:
title = 'Project {} Knowledge Graph'.format(self.data['name'])
args = {'title': title,
'node_properties': {},
'width': 2600,
'height': 2600,
'maxLinkWidth': 7,
'maxRadius': 20}
color_selector = "{'selector': '[name = \"KEY\"]', 'style': {'font-size': 10, 'background-color':'VALUE','width': 50,'height': 50,'background-image':'/assets/graph_icons/ENTITY.png','background-fit': 'cover','opacity':OPACITY}}"
stylesheet = [{'selector': 'node', 'style': {'label': 'data(name)', 'z-index': 9999}},
{'selector': 'edge', 'style': {'label': 'data(type)',
'curve-style': 'unbundled-bezier',
'control-point-distance': '20px',
'control-point-weight': '0.7',
'z-index': 5000,
'line-color': '#bdbdbd',
'opacity': 0.2,
'font-size': '7px'}}]
layout = {'name': 'circle'}
#stylesheet.extend([{'selector':'[weight < 0]', 'style':{'line-color':'#3288bd'}},{'selector':'[width > 0]', 'style':{'line-color':'#d73027'}}])
for n in self.nodes:
color = self.nodes[n]['color']
image = self.nodes[n]['type']
opacity = 0.3 if image == 'Module' or image == 'Group' else 1
stylesheet.append(ast.literal_eval(color_selector.replace("KEY", n.replace("'", "")).replace("VALUE", color).replace("ENTITY", image).replace("OPACITY", str(opacity))))
stylesheet.extend([{'selector': '[weight < 0]', 'style': {'line-color': '#4add1'}}, {'selector': '[weight > 0]', 'style': {'line-color': '#d6604d'}}])
args['stylesheet'] = stylesheet
args['layout'] = layout
nodes_table, edges_table = viz.network_to_tables(self.graph)
nodes_fig_table = viz.get_table(nodes_table, identifier=self.identifier+"_nodes_table", title="Nodes table")
edges_fig_table = viz.get_table(edges_table, identifier=self.identifier+"_edges_table", title="Edges table")
cy_elements, mouseover_node = utils.networkx_to_cytoscape(self.graph)
#args['mouseover_node'] = mouseover_node
net = {"notebook": [cy_elements, stylesheet, layout], "app": viz.get_cytoscape_network(cy_elements, self.identifier, args), "net_tables": (nodes_fig_table, edges_fig_table), "net_json": json_graph.node_link_data(self.graph)}
return net
def get_project_information(driver, project_id):
query_name = 'project_graph'
queries = []
data = []
res = []
try:
query = ''
data_upload_cypher = get_data_upload_queries()
for section in data_upload_cypher[query_name]:
code = section['query']
queries.extend(
code.replace("PROJECTID", project_id).split(';')[0:-1])
for query in queries:
result = connector.sendQuery(driver, query + ";")[0]
data.append(result)
except Exception as err:
exc_type, exc_obj, exc_tb = sys.exc_info()
fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
logger.error(
"Error: {}. Creating analytical samples: Query name ({}) - Query ({}), error info: {}, file: {},line: {}"
.format(err, query_name, query, sys.exc_info(), fname,
exc_tb.tb_lineno))
if data:
for i, j in enumerate(data):
df = pd.DataFrame([data[i]], columns=data[i].keys())
header = '_'.join(df.columns[0].split('_', 1)[1:]).capitalize()
df.rename(columns={df.columns[0]: 'project'}, inplace=True)
res.append(
viz.get_table(df,
identifier='new_project_{}'.format(header),
args={
'title':
'{} data uploaded for project {}'.format(
header, project_id)
}))
else:
res = None
logger.error(
"Error: No data was uploaded for project: {}. Review your experimental design and data files and the logs for errors."
.format(project_id))
return res
def generate_overlap_plots(self):
plots = []
identifier = "Overlap"
title = "Protein Identification Overlap"
plots.append(
viz.get_table(self.overlap, identifier + ' table',
title + ' table'))
if self.overlap is not None:
for i, row in self.overlap.iterrows():
ntitle = title + ":\n" + row['project1_name'] + " - " + row[
'project2_name'] + "(overlap similarity: " + str(
row['similarity']) + ")"
plot = viz.plot_2_venn_diagram(row['from'],
row['to'],
row['project1_unique'],
row['project2_unique'],
row['intersection'],
identifier=identifier + str(i),
args={'title': ntitle})
plots.append(plot)
return plots
def get_plot(self, name, identifier):
plot = []
if len(self.result) >= 1:
if name == "basicTable":
colors = ('#C2D4FF', '#F5F8FF')
columns = None
rows = None
figure_title = 'Basic table'
if "colors" in self.args:
colors = self.args["colors"]
if "cols" in self.args:
columns = self.args["cols"]
if "rows" in self.args:
rows = self.args["rows"]
if "title" in self.args:
figure_title = self.args["title"]
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args["title"] + id[
0] + " vs " + id[1]
plot.append(
viz.get_table(self.result[id],
identifier,
args={
'title': figure_title,
'colors': colors,
'cols': columns,
'rows': rows,
'width': 800,
'height': 1500,
'font': 12
}))
if name == "multiTable":
for id in self.result:
plot.append(
viz.get_multi_table(self.result[id], identifier,
self.args["title"]))
elif name == "barplot":
x_title = "x"
y_title = "y"
if "x_title" in self.args:
x_title = self.args["x_title"]
if "y_title" in self.args:
y_title = self.args["y_title"]
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args['title'] + id[
0] + " vs " + id[1]
else:
figure_title = self.args['title']
self.args["title"] = figure_title
plot.append(
viz.get_barplot(self.result[id], identifier,
self.args))
elif name == "facetplot":
x_title = "x"
y_title = "y"
plot_type = "bar"
if "x_title" not in self.args:
self.args["x_title"] = x_title
if "y_title" not in self.args:
self.args["y_title"] = y_title
if "plot_type" not in self.args:
self.args["plot_type"] = plot_type
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args['title'] + id[
0] + " vs " + id[1]
else:
figure_title = self.args['title']
self.args['title'] = figure_title
plot.append(
viz.get_facet_grid_plot(self.result[id], identifier,
self.args))
elif name == "scatterplot":
x_title = "x"
y_title = "y"
if "x_title" in self.args:
x_title = self.args["x_title"]
if "y_title" in self.args:
y_title = self.args["y_title"]
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args['title'] + id[
0] + " vs " + id[1]
else:
figure_title = self.args['title']
self.args['title'] = figure_title
plot.append(
viz.get_scatterplot(self.result[id], identifier,
self.args))
elif name == 'pca':
x_title = "x"
y_title = "y"
if "x_title" in self.args:
x_title = self.args["x_title"]
if "y_title" in self.args:
y_title = self.args["y_title"]
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args['title'] + id[
0] + " vs " + id[1]
else:
figure_title = self.args['title']
self.args['title'] = figure_title
plot.append(
viz.get_pca_plot(self.result[id], identifier,
self.args))
elif name == "volcanoplot":
alpha = 0.05
lfc = 1.0
if "alpha" not in self.args:
self.args["alpha"] = alpha
if "lfc" not in self.args:
self.args["lfc"] = lfc
for pair in self.result:
signature = self.result[pair]
self.args["title"] = self.args['title'] + " " + pair[
0] + " vs " + pair[1]
p = viz.run_volcano(
signature,
identifier + "_" + pair[0] + "_vs_" + pair[1],
self.args)
plot.extend(p)
elif name == 'network':
source = 'source'
target = 'target'
if "source" not in self.args:
self.args["source"] = source
if "target" not in self.args:
self.args["target"] = target
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args["title"] + id[
0] + " vs " + id[1]
else:
figure_title = self.args["title"]
self.args["title"] = figure_title
plot.append(
viz.get_network(self.result[id], identifier,
self.args))
elif name == "heatmap":
for id in self.result:
if not self.result[id].empty:
if isinstance(id, tuple):
identifier = identifier + "_" + id[
0] + "_vs_" + id[1]
figure_title = self.args["title"] + id[
0] + " vs " + id[1]
else:
figure_title = self.args["title"]
self.args["title"] = figure_title
plot.append(
viz.get_complex_heatmapplot(
self.result[id], identifier, self.args))
elif name == "mapper":
for id in self.result:
labels = {}
if "labels" not in self.args:
self.args["labels"] = labels
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args['title'] + id[
0] + " vs " + id[1]
else:
figure_title = self.args['title']
plot.append(
viz.getMapperFigure(self.result[id],
identifier,
title=figure_title,
labels=self.args["labels"]))
elif name == "scatterplot_matrix":
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args['title'] + id[
0] + " vs " + id[1]
else:
figure_title = self.args['title']
self.args["title"] = figure_title
plot.append(
viz.get_scatterplot_matrix(self.result[id], identifier,
self.args))
elif name == "distplot":
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args['title'] + id[
0] + " vs " + id[1]
else:
figure_title = self.args['title']
self.args["title"] = figure_title
plot.extend(
viz.get_distplot(self.result[id], identifier,
self.args))
elif name == "violinplot":
for id in self.result:
if isinstance(id, tuple):
identifier = identifier + "_" + id[0] + "_vs_" + id[1]
figure_title = self.args['title'] + id[
0] + " vs " + id[1]
else:
figure_title = self.args['title']
self.args["title"] = figure_title
plot.extend(
viz.get_violinplot(self.result[id], identifier,
self.args))
elif name == "polar":
for id in self.result:
figure_title = self.args['title']
plot.append(
viz.get_polar_plot(self.result[id], identifier,
self.args))
elif name == "km":
for id in self.result:
plot.append(
viz.get_km_plot(self.result[id], identifier,
self.args))
elif name == "wgcnaplots":
start = time.time()
data = {}
wgcna_data = self.result
if 'drop_cols_exp' in self.args and 'drop_cols_cli' in self.args:
if 'wgcna' in wgcna_data and wgcna_data[
'wgcna'] is not None:
for dtype in wgcna_data['wgcna']:
data = wgcna_data['wgcna'][dtype]
plot.extend(
viz.get_WGCNAPlots(data,
identifier + "-" + dtype))
print('WGCNA-plot', time.time() - start)
elif name == 'ranking':
for id in self.result:
plot.append(
viz.get_ranking_plot(self.result[id], identifier,
self.args))
elif name == 'qcmarkers_boxplot':
for id in self.result:
plot.append(
viz.get_boxplot_grid(self.result[id], identifier,
self.args))
elif name == 'clustergrammer':
for id in self.result:
plot.append(
viz.get_clustergrammer_plot(self.result[id],
identifier, self.args))
elif name == 'cytonet':
for id in self.result:
plot.append(
viz.get_cytoscape_network(self.result[id], identifier,
self.args))
elif name == 'wordcloud':
for id in self.result:
plot.append(
viz.get_wordcloud(self.result[id], identifier,
self.args))
self.update_plots({identifier: plot})
return plot
def get_knowledge_graph_plot(self, summarize=True):
if self.graph is None:
self.generate_knowledge_graph()
selected_nodes = []
if summarize and len(self.graph.nodes()) > 1:
centrality = nx.betweenness_centrality(self.graph,
k=None,
weight='weight',
normalized=False)
#centrality = nx.pagerank(G, alpha=0.95, weight='weight')
nx.set_node_attributes(self.graph, centrality, 'centrality')
sorted_centrality = sorted(centrality.items(),
key=itemgetter(1),
reverse=True)
for node_type in self.entities:
nodes = [
x for x, y in self.graph.nodes(data=True) if 'type' in y
and y['type'] == node_type and x not in self.keep_nodes
]
selected_nodes.extend(
[n for n, c in sorted_centrality if n in nodes][15:])
if len(selected_nodes) > 0:
self.graph.remove_nodes_from(selected_nodes)
title = 'Project {} Knowledge Graph'.format(self.identifier)
if self.data is not None:
if 'name' in self.data:
title = 'Project {} Knowledge Graph'.format(self.data['name'])
args = {
'title': title,
'node_properties': {},
'width': 2000,
'height': 2000,
'maxLinkWidth': 7,
'maxRadius': 20
}
color_selector = "{'selector': '[name = \"KEY\"]', 'style': {'font-size': '7px', 'text-opacity': 0.8, 'background-color':'VALUE','width': 50,'height': 50,'background-image':'/assets/graph_icons/ENTITY.png','background-fit': 'cover','opacity':OPACITY}}"
stylesheet = [{
'selector': 'node',
'style': {
'label': 'data(name)',
'opacity': 0.7
}
}, {
'selector': 'edge',
'style': {
'label': 'data(type)',
'curve-style': 'unbundled-bezier',
'control-point-distance': '30px',
'control-point-weight': '0.7',
'z-index': 5000,
'line-color': '#bdbdbd',
'opacity': 0.2,
'font-size': '2.5px',
'text-opacity': 1,
'font-style': "normal",
'font-weight': "normal"
}
}]
layout = {
'name': 'cose',
'idealEdgeLength': 100,
'nodeOverlap': 20,
'refresh': 20,
'randomize': False,
'componentSpacing': 100,
'nodeRepulsion': 400000,
'edgeElasticity': 100,
'nestingFactor': 5,
'gravity': 80,
'numIter': 1000,
'initialTemp': 200,
'coolingFactor': 0.95,
'minTemp': 1.0
}
stylesheet.extend([{
'selector': '[weight < 0]',
'style': {
'line-color': '#3288bd'
}
}, {
'selector': '[weight > 0]',
'style': {
'line-color': '#d73027'
}
}])
for n, attr in self.graph.nodes(data=True):
color = self.default_color
image = ''
if 'color' in attr:
color = attr['color']
if 'type' in attr:
image = attr['type']
opacity = 0.3 if image == 'Module' or image == 'Group' else 1
stylesheet.append(
ast.literal_eval(
color_selector.replace("KEY", n.replace("'", "")).replace(
"VALUE",
color).replace("ENTITY",
image).replace("OPACITY",
str(opacity))))
stylesheet.extend([{
'selector': 'node',
'style': {
'width': 'mapData(centrality, 0, 1, 15, 30)',
'height': 'mapData(centrality, 0, 1, 15, 30)'
}
}])
args['stylesheet'] = stylesheet
args['layout'] = layout
G = self.graph.copy()
if G.has_node('Regulated'):
G.remove_node('Regulated')
nodes_table, edges_table = viz.network_to_tables(G,
source='node1',
target='node2')
nodes_fig_table = viz.get_table(nodes_table,
identifier=self.identifier +
"_nodes_table",
args={'title': "Nodes table"})
edges_fig_table = viz.get_table(edges_table,
identifier=self.identifier +
"_edges_table",
args={'title': "Edges table"})
cy_elements, mouseover_node = utils.networkx_to_cytoscape(G)
#args['mouseover_node'] = mouseover_node
net = {
"notebook": [cy_elements, stylesheet, layout],
"app": viz.get_cytoscape_network(cy_elements, self.identifier,
args),
"net_tables": (nodes_table, edges_table),
"net_tables_viz": (nodes_fig_table, edges_fig_table),
"net_json": json_graph.node_link_data(G)
}
return net
