Plot Sequence Logos with negative heights

Question

I am plotting text with TextPath from the matplotlib.text module. I partially use the code from this thread and plot also with negative heights. It works quite well except for the fact that code plotted with negative height is flipped vertically. While this is neither a problem for "A", nor "T". "C" and especially "G" look very strange:

Is there a way to actually flip the text vertically? Here is part of the code, although mostly copied from the thread mentioned above

import matplotlib as mpl
from matplotlib.text import TextPath
from matplotlib.patches import PathPatch
from matplotlib.font_manager import FontProperties
import matplotlib.pyplot as plt
import itertools
import operator
import numpy as np

fp = FontProperties(family="Arial", weight="bold") 
globscale = 1.35
LETTERS = { "T" : TextPath((-0.305, 0), "T", size=1, prop=fp),
            "G" : TextPath((-0.384, 0), "G", size=1, prop=fp),
            "A" : TextPath((-0.35, 0), "A", size=1, prop=fp),
            "C" : TextPath((-0.366, 0), "C", size=1, prop=fp) }
COLOR_SCHEME = {'G': 'gold', 
                'A': 'forestgreen', 
                'C': 'mediumblue', 
                'T': 'crimson'}

def letterAt(letter, x, y, yscale=1, ax=None):
    text = LETTERS[letter]

    t = mpl.transforms.Affine2D().scale(1*globscale, yscale*globscale) + \
        mpl.transforms.Affine2D().translate(x,y) + ax.transData
    p = PathPatch(text, lw=0, fc=COLOR_SCHEME[letter],  transform=t)
    if ax != None:
        ax.add_artist(p)
    return p

ALL_SCORES1 = [[('C', 0.02247014831444764),
          ('T', 0.057903843733384308),
          ('A', 0.10370837683591219),
          ('G', 0.24803586793255664)],
         [('T', 0.046608227674354567),
          ('G', 0.048827667087419063),
          ('A', 0.084338697696451109),
          ('C', 0.92994511407402669)],
         [('G', 0.0),
          ('T', 0.011098351287382456),
          ('A', 0.022196702574764911),
          ('C', 1.8164301607015951)],
         [('C', 0.020803153636453006),
          ('T', 0.078011826136698756),
          ('G', 0.11268374886412044),
          ('A', 0.65529933954826969)],
         [('T', 0.017393530660176126),
          ('A', 0.030438678655308221),
          ('G', 0.22611589858228964),
          ('C', 0.45078233627623127)],
         [('G', 0.022364103549245576),
          ('A', 0.043412671595594352),
          ('T', 0.097349627214363091),
          ('C', 0.1657574733649966)],
         [('C', 0.03264675899941203),
          ('T', 0.045203204768416654),
          ('G', 0.082872542075430544),
          ('A', 1.0949220710572034)],
         [('C', 0.0),
          ('T', 0.0076232429756614498),
          ('A', 0.011434864463492175),
          ('G', 1.8867526364762088)],
         [('C', 0.0018955903000026028),
          ('T', 0.0094779515000130137),
          ('A', 0.35637097640048931),
          ('G', 0.58005063180079641)],
         [('A', 0.01594690817903021),
          ('C', 0.017541598996933229),
          ('T', 0.2774762023151256),
          ('G', 0.48638069946042134)],
         [('A', 0.003770051401807444),
          ('C', 0.0075401028036148881),
          ('T', 0.011310154205422331),
          ('G', 1.8624053924928772)],
         [('C', 0.036479877757360731),
          ('A', 0.041691288865555121),
          ('T', 0.072959755514721461),
          ('G', 1.1517218549109602)],
         [('G', 0.011831087684038642),
          ('T', 0.068620308567424126),
          ('A', 0.10174735408273231),
          ('C', 1.0009100180696691)],
         [('C', 0.015871770937774379),
          ('T', 0.018757547471915176),
          ('A', 0.32176408355669878),
          ('G', 0.36505073156881074)],
         [('A', 0.022798100897300954),
          ('T', 0.024064662058262118),
          ('G', 0.24571286522646588),
          ('C', 0.34070495229855319)]]


fig, ax = plt.subplots(figsize=(10,3))

all_scores = ALL_SCORES1
x = 1
maxi = 2
for scores in all_scores:
    y = 0
    for base, score in scores:
        letterAt(base, x,y, -score, ax)
        y -= score
    x += 1
    maxi = max(maxi, y)

plt.xticks(range(1,x))
plt.xlim((0, x)) 
plt.ylim((-2, maxi)) 
plt.tight_layout()      
plt.show()

I tried a lot but couldn't find an answer. look at [this](https://stackoverflow.com/questions/9154120/flipping-a-picture-across-the-vertical-axis-in-python) it may be helpful. — Krishna, Jul 09 '18 at 09:10
From the image you show, the text is flipped vertically, not horizontally. Can you be more clear on what you want to achieve? Why not plot all letters in a non-flipped fashion? — ImportanceOfBeingErnest, Jul 09 '18 at 09:13
@ImportanceOfBeingErnest you are completly right! I should not flip them but just plot them with an offset. thank you! — MaxS, Jul 09 '18 at 09:26

score 1 · Answer 1 · answered Jul 09 '18 at 11:04

ImportanceOfBeingErnest pointed me towards the solution: It is neither a flipping of the letters nor a plotting with negative heights but just a matter of changing the y-value, the plot starts at:

fig, ax = plt.subplots(figsize=(10,3))
x = 1.0
maxi = 2
for scores in all_scores:
#sort the scores with "reverse=True" so most important letters are at the bottom
    scores=sorted(scores,key=operator.itemgetter(1),reverse=True)
#start plotting at the negative sum of all values below 0
    y = np.sum([-1*s[1] for s in scores])
    for base, score in scores:
        letterAt(base, x,y, score, ax)
        y += score
    x += 1
    maxi = max(maxi, y)


plt.xticks(np.arange(1,x))
plt.xlim((0, x)) 
plt.ylim((-2, maxi)) 
plt.tight_layout()      
plt.show()

@ImportanceOfBeingErnest Would you object in any way if I use your answer from the thread "sequence logos in matplotlib: aligning xticks" to publish my script on GitHub? I used it also to plot base inter dependencies and to my knowledge, there is nothing like this in python available and it might save people some work. — MaxS, Jul 10 '18 at 11:53

Plot Sequence Logos with negative heights

1 Answers1