def getHeatmap(filename):
try:
a = data(filename).getMatrix()
except KeyError:
a = loadTSV(filename)
heatmap(a)
plt.show()
def getTsne(filepath, targetname='', cluster_label=[]):
try:
data = loadTSV(filepath)
except:
data = loadCSV(filepath)
X_tsne = TSNE(learning_rate=200, random_state=0).fit_transform(data)
Xtsne = X_tsne[:, 0]
Ytsne = X_tsne[:, 1]
lisXtsne = Xtsne.tolist()
lisYtsne = Ytsne.tolist()
if not targetname == '':
colors = [
'black', 'purple', 'red', 'yellow', 'pink', 'grey', 'green',
'bisque', 'blanchedalmond', 'blue', 'blueviolet', 'brown',
'burlywood', 'cadetblue', 'chartreuse', 'chocolate'
]
if len(cluster_label) == 0:
cluster_label = [0] * len(lisXtsne)
fig = plt.figure(figsize=(20, 10),
dpi=80,
facecolor='w',
edgecolor='k')
for i in range(len(lisXtsne)):
plt.scatter(lisXtsne[i], lisYtsne[i], c=colors[cluster_label[i]])
plt.title(targetname)
plt.xlabel('x')
plt.ylabel('y')
fig.savefig(targetname)
return (lisXtsne, lisYtsne)
def getGeneofCentroids(filename, target, index):
# save gene lists of centroid cells
matrix = loadTSV(filename)
result = np.zeros(shape=(len(index), len(matrix[0])))
num = 0
for i in index:
result[num, :] = matrix[i, :]
np.savetxt(target, result, delimiter="\t")
def update_graph1(visualization, dim_method, gene, dataset):
# load
if not visualization:
filepath = './data/' + dataset + "/" + dim_method + '/tsne.txt'
tsne = load(filepath)
else:
filepath = './data/' + dataset + "/" + dim_method + '/scvis.tsv'
scvis = loadTSV(filepath)
tsne = (scvis[:, 1], scvis[:, 2])
# generate range
rx = [min(tsne[0]) - 10, max(tsne[0]) + 10]
ry = [min(tsne[1]) - 10, max(tsne[1]) + 10]
return {
'data': [
go.Scattergl(
x=tsne[0],
y=tsne[1],
mode="markers",
marker=dict(
# set color equal to a variable
color=color_mask_genes[gene],
colorscale='Jet',
showscale=True))
],
'layout':
go.Layout(autosize=False,
title="TSNE with Genes",
font={
"family": "Raleway",
"size": "5vh"
},
height="200vh",
width="30%",
hovermode="closest",
margin={
"r": 0,
"t": 50,
"b": 40,
"l": 100
},
xaxis={
"range": rx,
"zeroline": False
},
yaxis={
"mirror": False,
"range": ry,
"zeroline": False,
})
}
def kmeans(filename, target_name="", clusters=8, scvis=False):
# use kmeans on tsne
if scvis:
matrix = loadTSV(filename)
else:
matrix = tsne(filename)
# get kmeans labels
k = KMeans(n_clusters=clusters).fit_predict(matrix)
# generate graph
if not target_name=="":
getTsne(filename, target_name, k)
return KMeans(n_clusters=clusters), k
def saveGeneTable(filename, file_after_reduction, target_dir, k, scvis=False):
createDir(target_dir)
_, k_mask = cluster.kmeans(file_after_reduction, clusters=k, scvis=scvis)
data = loadTSV(filename)
indexes = []
for _ in range(k):
indexes.append([])
for index in range(len(k_mask)):
indexes[k_mask[index]].append(index)
result = np.zeros(shape=(k, len(data[0])))
for _ in range(k):
result[_] = np.mean(data[indexes[_]], axis=0)
if not scvis:
np.savetxt(target_dir + "geneTable_" + str(k) + ".txt",
result.transpose(),
delimiter="\t")
else:
np.savetxt(target_dir + "geneTable_" + str(k) + "_scvis.txt",
result.transpose(),
delimiter="\t")
result[i, j] = float(lines[i].split()[j])
return result
def loadList(file):
with open(file, "r") as o:
data = o.read()
lines = data.splitlines()
result = []
for n_ in range(len(lines)):
result.append(lines[n_])
return result
# load genelist
genelist = loadList("./data/Airway/genelist.txt")
filtered = loadTSV("./data/Airway/filtered.txt")
dropdown_label = []
color_mask_genes = []
for i in range(len(genelist)):
dropdown_label.append({"label": genelist[i], "value": i})
color_mask_genes.append(filtered[:, i])
# draw website
app = dash.Dash()
app.layout = html.Div(style={"height": "150vh", "fontFamily": "Georgia"}, children=[
# head
html.Div(style={'background-color': "#CEF0EF", "height": "10vh", "width": "100%"}, children=[
html.Img(src='https://upload.wikimedia.org/wikipedia/commons/e/e1/Duke_Athletics_logo.svg', height="50vh",
width="50vh", style={"margin-top": "2vh", "margin-left": "3vh"}),
html.Ul(style={"float": "right", "list-style": "none", "margin-right": "5vh"},
def tsne(filepath):
# get tsne coordinates
return TSNE(random_state=0).fit_transform(loadTSV(filepath))
def savePCA(file, target_dir):
createDir(target_dir)
data = loadTSV(file)
pca = PCA(n_components=10).fit_transform(data)
print(type(pca))
np.savetxt(target_dir + "pca.txt", pca, delimiter="\t")
def getData(filepath):
data = data_helper.loadTSV(filepath)
return data
def update_graph_2(value, dim_method, dataset, visualization):
filepath = './data/' + dataset + "/" + dim_method + "/"
# load
if not visualization:
filepath += "tsne.txt"
tsne = load(filepath)
else:
filepath = './data/' + dataset + "/" + dim_method + '/scvis.tsv'
scvis = loadTSV(filepath)
tsne = (scvis[:, 1], scvis[:, 2])
# load colormap
color_mask = []
if not visualization:
for k_ in range(1, 9):
color_mask.append(
loadColorMask("./data/" + dataset + "/" + dim_method + "/" +
"color_mask_" + str(k_) + ".txt"))
else:
for k_ in range(1, 9):
color_mask.append(
loadColorMask("./data/" + dataset + "/" + dim_method + "/" +
"color_mask_" + str(k_) + "_scvis.txt"))
text = []
for i in color_mask[value - 1]:
text.append("Cluster " + str(i))
rx = [min(tsne[0]) - 10, max(tsne[0]) + 10]
ry = [min(tsne[1]) - 10, max(tsne[1]) + 10]
return {
'data': [
go.Scattergl(
x=tsne[0],
y=tsne[1],
text=text,
mode="markers",
marker=dict(
color=color_mask[value -
1], # set color equal to a variable
showscale=True,
colorscale="Jet"))
],
'layout':
go.Layout(autosize=False,
title="TSNE with KMeans",
font={
"family": "Raleway",
"size": "5vh"
},
height="200vh",
width="40%",
hovermode="closest",
margin={
"r": 0,
"t": 50,
"b": 40,
"l": 100
},
xaxis={
"range": rx,
"zeroline": False
},
yaxis={
"range": ry,
"zeroline": False
})
}