AKAGI/logo.py at master · farzinlize/AKAGI · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
from math import log2, log
from TrieFind import ChainNode
from alignment import alignment_strings
from logomaker import Logo, alignment_to_matrix
from pandas import DataFrame
from misc import maxSubArraySum, make_location
import matplotlib.pyplot as plt

from constants import PSEUDOCOUNT

def make_pattern_report(name, pattern:ChainNode, sequences, directory=''):

    def no_negative(data_frame):
        for alphabet in 'ATCG':
            for i in range(len(data_frame)):
                if data_frame[alphabet][i] < 0:data_frame[alphabet][i] = 0

    make_location(directory)
    cal = lambda col, row, count:log2( ( (col[row]+(PSEUDOCOUNT/4) ) / (count+PSEUDOCOUNT) )/0.25 )

    binding_sites = pattern.instances_list(sequences)
    aligned_sites = alignment_strings(binding_sites)
    aligned_matrix = alignment_to_matrix(aligned_sites)

    position_counts = [sum(aligned_matrix[alphabet][i] for alphabet in 'ACGT') for i in range(len(aligned_matrix))]
    average_count = sum(position_counts) / len(position_counts)
    shifted_counts = [c-average_count for c in position_counts]
    log_shifted_counts = [log(c/average_count) for c in position_counts]

    _, start, end         = maxSubArraySum(shifted_counts)
    _, log_start, log_end = maxSubArraySum(log_shifted_counts)

    plt.plot(position_counts)
    plt.ylabel('position counts')
    plt.xlabel('alignment positions')
    plt.savefig(f'{directory}{name}_bs_alignment_positioncount.png')
    plt.clf()

    plt.plot(shifted_counts)
    plt.ylabel('shifted position counts by average')
    plt.xlabel('alignment positions')
    plt.axhline(0, color='red')
    plt.axvline(start, color='red')
    plt.axvline(end, color='red')
    plt.savefig(f'{directory}{name}_bs_alignment_positioncount_shifted.png')
    plt.clf()

    plt.plot(log_shifted_counts)
    plt.ylabel('log shifted position counts by average')
    plt.xlabel('alignment positions')
    plt.axhline(0, color='red')
    plt.axvline(log_start, color='red')
    plt.axvline(log_end, color='red')
    plt.savefig(f'{directory}{name}_bs_alignment_positioncount_logshifted.png')
    plt.clf()

    alignment_logo = Logo(aligned_matrix)
    alignment_logo.draw()
    plt.xlabel('alignment by muscle')
    plt.savefig(f'{directory}{name}_bs_alignment.png')
    plt.clf()

    A = list(aligned_matrix['A'][start:end+1])
    C = list(aligned_matrix['C'][start:end+1])
    T = list(aligned_matrix['T'][start:end+1])
    G = list(aligned_matrix['G'][start:end+1])

    Al = list(aligned_matrix['A'][log_start:log_end+1])
    Cl = list(aligned_matrix['C'][log_start:log_end+1])
    Tl = list(aligned_matrix['T'][log_start:log_end+1])
    Gl = list(aligned_matrix['G'][log_start:log_end+1])

    # r_start     = len(aligned_matrix) - end
    # r_end       = len(aligned_matrix) - start
    # r_log_start = len(aligned_matrix) - log_end
    # r_log_end   = len(aligned_matrix) - log_start

    rA = list(reversed(list(aligned_matrix['T'][start:end+1])))
    rC = list(reversed(list(aligned_matrix['G'][start:end+1])))
    rG = list(reversed(list(aligned_matrix['C'][start:end+1])))
    rT = list(reversed(list(aligned_matrix['A'][start:end+1])))

    rAl = list(reversed(list(aligned_matrix['T'][log_start:log_end+1])))
    rCl = list(reversed(list(aligned_matrix['G'][log_start:log_end+1])))
    rGl = list(reversed(list(aligned_matrix['C'][log_start:log_end+1])))
    rTl = list(reversed(list(aligned_matrix['A'][log_start:log_end+1])))

    data = {'A':[cal(A, i, A[i]+T[i]+C[i]+G[i]) for i in range(len(A))],
            'T':[cal(T, i, A[i]+T[i]+C[i]+G[i]) for i in range(len(T))],
            'C':[cal(C, i, A[i]+T[i]+C[i]+G[i]) for i in range(len(C))],
            'G':[cal(G, i, A[i]+T[i]+C[i]+G[i]) for i in range(len(G))]}

    rdata = {'A':[cal(rA, i, rA[i]+rT[i]+rC[i]+rG[i]) for i in range(len(rA))],
             'T':[cal(rT, i, rA[i]+rT[i]+rC[i]+rG[i]) for i in range(len(rT))],
             'C':[cal(rC, i, rA[i]+rT[i]+rC[i]+rG[i]) for i in range(len(rC))],
             'G':[cal(rG, i, rA[i]+rT[i]+rC[i]+rG[i]) for i in range(len(rG))]}

    datal = {'A':[cal(Al, i, Al[i]+Tl[i]+Cl[i]+Gl[i]) for i in range(len(Al))],
             'T':[cal(Tl, i, Al[i]+Tl[i]+Cl[i]+Gl[i]) for i in range(len(Tl))],
             'C':[cal(Cl, i, Al[i]+Tl[i]+Cl[i]+Gl[i]) for i in range(len(Cl))],
             'G':[cal(Gl, i, Al[i]+Tl[i]+Cl[i]+Gl[i]) for i in range(len(Gl))]}

    rdatal = {'A':[cal(rAl, i, rAl[i]+rTl[i]+rCl[i]+rGl[i]) for i in range(len(rAl))],
              'T':[cal(rTl, i, rAl[i]+rTl[i]+rCl[i]+rGl[i]) for i in range(len(rTl))],
              'C':[cal(rCl, i, rAl[i]+rTl[i]+rCl[i]+rGl[i]) for i in range(len(rCl))],
              'G':[cal(rGl, i, rAl[i]+rTl[i]+rCl[i]+rGl[i]) for i in range(len(rGl))]}

    df   = DataFrame(data)  ;no_negative(df)
    rdf  = DataFrame(rdata) ;no_negative(rdf)
    dfl  = DataFrame(datal) ;no_negative(dfl)
    rdfl = DataFrame(rdatal);no_negative(rdfl)

    pwm_logo = Logo(df)
    pwm_logo.draw()
    plt.savefig(f'{directory}{name}_bs_shorten_pwm.png')
    plt.clf()

    rpwm_logo = Logo(rdf)
    rpwm_logo.draw()
    plt.savefig(f'{directory}{name}_bs_shorten_rpwm.png')
    plt.clf()

    pwml_logo  = Logo(dfl)
    pwml_logo.draw()
    plt.savefig(f'{directory}{name}_bs_shorten_pwml.png')
    plt.clf()

    rpwml_logo = Logo(rdfl)
    rpwml_logo.draw()
    plt.savefig(f'{directory}{name}_bs_shorten_rpwml.png')
    plt.clf()