cvmcore 0.1.1

SZQ lab data analysis core function

cvmcore

Introduction

The core function of data analysis for plot or data process used by SZQ lab from China Agricultural University

Example usage

from Bio import Phylo
import matplotlib as mpl
import matplotlib.pyplot as plt
from io import StringIO
import matplotlib.collections as mpcollections
from copy import copy

import pandas as pd
import numpy as np
import seaborn as sn

from cvmcore.cvmcore import cvmplot

from scipy.cluster.hierarchy import linkage, dendrogram, complete, to_tree
from scipy.spatial.distance import squareform

mlst = [[np.nan, 19., 12.,  9.,  5.,  9.,  2.],
        [np.nan, 19., 12.,  9.,  5.,  9.,  2.],
        [10., 17., 12.,  9., np.nan,  9.,  2.],
        [10., 19., 12., np.nan,  5.,  9.,  2.],
        [np.nan, 19., 13.,  9.,  5.,  9.,  2.]]
genes = np.char.replace(np.array(np.arange(1, 8), dtype='str'), '', 'gene_', count=1)
samples = np.char.replace(np.array(np.arange(1, 6), dtype='str'), '', 'sample_', count=1)
df_mlst = pd.DataFrame(mlst, index=samples, columns=genes)
diff_matrix = cvmplot.get_diff_df(df_mlst)

diff_matrix

	sample_1	sample_2	sample_3	sample_4	sample_5
sample_1	0	0	1	0	1
sample_2	0	0	1	0	1
sample_3	1	1	0	1	2
sample_4	0	0	1	0	1
sample_5	1	1	2	1	0

link_matrix =linkage(squareform(diff_matrix), method='complete')

link_matrix

array([[0., 1., 0., 2.],
       [3., 5., 0., 3.],
       [2., 6., 1., 4.],
       [4., 7., 2., 5.]])

1. Plot a rectangular dendrogram

fig, ax= plt.subplots(1,1)
lableorder, ax = cvmplot.rectree(link_matrix, scale_max=7, labels=samples, ax=ax)
fig.tight_layout()
fig.savefig('Screenshots/dendrogram.png')

2. Plot rectangular dendrogram with heatmap

#create dataframe
mat = np.random.randint(70, 100, (5, 10))
loci = np.char.replace(np.array(np.arange(1, 11), dtype='str'), '', 'loci_', count=1)
sample = np.char.replace(np.array(np.arange(1, 6), dtype='str'), '', 'sample', count=1)
df_heatmap = pd.DataFrame(mat, index=sample, columns=loci)

#create linkage matrix
diff_matrix = [[0, 0, 1, 0, 1],
               [0, 0, 1, 0, 1],
               [1, 1, 0, 1, 2],
               [0, 0, 1, 0, 1],
               [1, 1, 2, 1, 0]]

linkage_matrix = linkage(squareform(diff_matrix),'complete')

fig, (ax1, ax2) = plt.subplots(1,2,figsize=(8,3), gridspec_kw={'width_ratios': [1, 2]})
fig.tight_layout(w_pad=-2)

row_order, ax1 = cvmplot.rectree(linkage_matrix,labels=sample, no_labels=True, scale_max=3, ax=ax1)
cvmplot.heatmap(df_heatmap, order=row_order, ax=ax2, cbar=True)

ax1.set_xticklabels(ax1.get_xticklabels(), fontsize=15)

ax2.set_xticklabels(ax2.get_xticklabels(), rotation=90, fontsize=15)
ax2.set_yticklabels(ax2.get_yticklabels(), fontsize=15)
ax2.xaxis.tick_top()

fig.savefig('Screenshots/test.pdf', bbox_inches='tight')

[ 5 15 25 35 45]
['sample5', 'sample3', 'sample4', 'sample1', 'sample2']

fig, (ax1, ax2, ax3) = plt.subplots(1,3,figsize=(12,3), gridspec_kw={'width_ratios': [1, 2, 2]})

fig.tight_layout(w_pad=-2)

row_order, ax1 = cvmplot.rectree(linkage_matrix,labels=sample, no_labels=True, scale_max=3, ax=ax1)

# remove the yticklabels in ax2
ax2 = cvmplot.heatmap(df_heatmap, order=row_order, ax=ax2, cbar=True, yticklabel=False)
# add ax3 heatmap
ax3 = cvmplot.heatmap(df_heatmap, order=row_order, ax=ax3, cmap='Blues', cbar=True, yticklabel=True)

#set ticklabels property of x or y from ax1, ax2, ax3
ax1.set_xticklabels(ax1.get_xticklabels(), fontsize=15)

ax2.set_xticklabels(ax2.get_xticklabels(), rotation=90, fontsize=15)
ax2.xaxis.tick_top()

ax3.set_xticklabels(ax3.get_xticklabels(), rotation=90, fontsize=15)
ax3.set_yticklabels(ax3.get_yticklabels(), fontsize=15)
ax3.xaxis.tick_top()


# fig.tight_layout()
fig.savefig('Screenshots/multiple_heatmap.png', bbox_inches='tight')

2.1 set minimum value of heatmap

fig, (ax1, ax2) = plt.subplots(1,2,figsize=(8,3), gridspec_kw={'width_ratios': [1, 2]})
fig.tight_layout(w_pad=-2)

order, ax1 = cvmplot.rectree(linkage_matrix,labels=sample, no_labels=True, scale_max=3, ax=ax1)
cvmplot.heatmap(df_heatmap, order=order, ax=ax2, cbar=True, vmin=90)

ax1.set_xticklabels(ax1.get_xticklabels(), fontsize=15)

ax2.set_xticklabels(ax2.get_xticklabels(), rotation=90, fontsize=15)
ax2.set_yticklabels(ax2.get_yticklabels(), fontsize=15)
ax2.xaxis.tick_top()

fig.savefig('Screenshots/dendrogram_heatmap_minimumvalue.pdf', bbox_inches='tight')

[ 5 15 25 35 45]
['sample5', 'sample3', 'sample4', 'sample1', 'sample2']

2.2 using cmap to change color

fig, (ax1, ax2) = plt.subplots(1,2,figsize=(8,3), gridspec_kw={'width_ratios': [1, 2]})
fig.tight_layout(w_pad=-2)

order, ax1 = cvmplot.rectree(linkage_matrix,labels=sample, no_labels=True, scale_max=3, ax=ax1)
cvmplot.heatmap(df_heatmap, order=order, ax=ax2, cmap='tab20', cbar=True)

ax1.set_xticklabels(ax1.get_xticklabels(), fontsize=15)

ax2.set_xticklabels(ax2.get_xticklabels(), rotation=90, fontsize=15)
ax2.set_yticklabels(ax2.get_yticklabels(), fontsize=15)
ax2.xaxis.tick_top()
fig.savefig('Screenshots/dendrogram_heatmap_cmap.pdf', bbox_inches='tight')

[ 5 15 25 35 45]
['sample5', 'sample3', 'sample4', 'sample1', 'sample2']

3. Plot a circular dendrogram

# generate two clusters: a with 100 points, b with 50:
np.random.seed(4711)  # for repeatability of this tutorial
a = np.random.multivariate_normal([10, 0], [[3, 1], [1, 4]], size=[100,])
b = np.random.multivariate_normal([0, 20], [[3, 1], [1, 4]], size=[50,])
X = np.concatenate((a, b),)

Z = linkage(X, 'ward')

Z2 = dendrogram(Z, no_plot=True)

# set open angle
fig, ax= plt.subplots(1,1,figsize=(10,10))

cvmplot.circulartree(Z2,addlabels=True, fontsize=10, ax=ax)
fig.tight_layout()
fig.savefig('Screenshots/circular_dendrogram.png', bbox_inches='tight')

3.1 color label

colors = [{'#0070c7':'2021'}, {'#3a9245':'2022'}, {'#f8d438':'2023'}]
result = np.random.choice(colors, size=150)
label_colors_map = dict(zip(Z2['ivl'], result))
point_colors_map = dict(zip(Z2['ivl'], result))

fig, ax= plt.subplots(1,1,figsize=(10,10))
cvmplot.circulartree(Z2, addlabels=True, branch_color=False, label_colors= label_colors_map, fontsize=15)
fig.tight_layout()
fig.savefig('Screenshots/circular_dendrogram_color_label.png')

3.2 set open angle

fig, ax= plt.subplots(1,1,figsize=(10,10))
cvmplot.circulartree(Z2, addlabels=True, branch_color=False, label_colors= label_colors_map, fontsize=15, open_angle=30)
fig.tight_layout()
fig.savefig('Screenshots/circular_dendrogram_openangle.png')

3.3 set start angle

fig, ax= plt.subplots(1,1,figsize=(10,10))
cvmplot.circulartree(Z2, addlabels=True, branch_color=False, label_colors= label_colors_map, fontsize=15, open_angle=90,
                     start_angle=30
                    )
fig.tight_layout()
fig.savefig('Screenshots/circular_dendrogram_startangle.png')

3.4 add point

fig, ax= plt.subplots(1,1,figsize=(12,10))
cvmplot.circulartree(Z2, addlabels=True, branch_color=False, label_colors= label_colors_map, fontsize=15, addpoints=True,
                     point_colors = point_colors_map, point_legend_title='Species', pointsize=25)
fig.tight_layout()
fig.savefig('Screenshots/circular_dendrogram_tippoints.png')

4. Plot phylogenetic tree

tree = "(((A:0.2, B:0.3):0.3,(C:0.5, D:0.3):0.2):0.3, E:0.7):1.0;"
tree = Phylo.read(StringIO(tree), 'newick')

fig, ax= plt.subplots(1,1, figsize=(10, 10))
ax, lable_order = cvmplot.phylotree(tree=tree, color='k', lw=1, ax=ax, show_label=True, align_label=True, labelsize=15)
fig.tight_layout()
fig.savefig('Screenshots/phylogenetic tree.png')

4.1 Plot tree with heatmap

#create dataframe
mat = np.random.randint(70, 100, (5, 10))
col = np.char.replace(np.array(np.arange(1, 11), dtype='str'), '', 'column_', count=1)
strains = ['A', 'B', 'C', 'D', 'E']
df_heatmap = pd.DataFrame(mat, index=strains, columns=col)

df_heatmap

	column_1	column_2	column_3	column_4	column_5	column_6	column_7	column_8	column_9	column_10
A	89	73	91	75	95	90	93	74	99	97
B	73	90	75	89	85	72	82	85	96	82
C	84	82	86	74	72	75	91	83	97	98
D	72	77	72	98	79	73	87	91	98	94
E	88	75	88	73	77	72	74	73	99	86

fig,(ax1, ax2)= plt.subplots(1,2, figsize=(8, 3), gridspec_kw={'width_ratios':[1, 2]})
fig.tight_layout(w_pad=-2)
ax1, order = cvmplot.phylotree(tree=tree, color='k', lw=1, ax=ax1, show_label=True, align_label=True, labelsize=15)
cvmplot.heatmap(df_heatmap, order=order, ax=ax2, cbar=True, vmin=90)

ax1.set_xticklabels(ax1.get_xticklabels(), fontsize=15)

ax2.set_xticklabels(ax2.get_xticklabels(), rotation=90, fontsize=15)
ax2.set_yticklabels(ax2.get_yticklabels(), fontsize=15)
ax2.xaxis.tick_top()

fig.savefig('Screenshots/phylotree_with_heatmap.pdf')

4.2 remove labels at the tip of the tree

fig,(ax1, ax2)= plt.subplots(1,2, figsize=(8, 3), gridspec_kw={'width_ratios':[1, 2]})
fig.tight_layout(w_pad=-2)
ax1, order = cvmplot.phylotree(tree=tree, color='k', lw=1, ax=ax1, show_label=False)
cvmplot.heatmap(df_heatmap, order=order, ax=ax2, cbar=True, vmin=90)

ax1.set_xticklabels(ax1.get_xticklabels(), fontsize=15)

ax2.set_xticklabels(ax2.get_xticklabels(), rotation=90, fontsize=15)
ax2.set_yticklabels(ax2.get_yticklabels(), fontsize=15)
ax2.xaxis.tick_top()

fig.savefig('Screenshots/phylotree_with_heatmap-remove_tiplable.pdf', bbox_inches='tight')

4.3 Plot multiple heatmap with phylotree

fig,(ax1, ax2, ax3)= plt.subplots(1,3, figsize=(12, 3), gridspec_kw={'width_ratios':[1, 2, 2]})
fig.tight_layout(w_pad=-2)
ax1, order = cvmplot.phylotree(tree=tree, color='k', lw=1, ax=ax1, show_label=True, align_label=True, labelsize=15)
ax2 = cvmplot.heatmap(df_heatmap, order=order, ax=ax2, cbar=True, vmin=90, yticklabel=False)
# add ax3 heatmap
ax3 = cvmplot.heatmap(df_heatmap, order=order, ax=ax3, cmap='Blues', cbar=True, yticklabel=True)

#set ticklabels property of x or y from ax1, ax2, ax3
ax1.set_xticklabels(ax1.get_xticklabels(), fontsize=15)

ax2.set_xticklabels(ax2.get_xticklabels(), rotation=90, fontsize=15)
ax2.xaxis.tick_top()

ax3.set_xticklabels(ax3.get_xticklabels(), rotation=90, fontsize=15)
ax3.set_yticklabels(ax3.get_yticklabels(), fontsize=15)
ax3.xaxis.tick_top()


# fig.tight_layout()
fig.savefig('Screenshots/phylotree_multiple_heatmap.png', bbox_inches='tight')