plot

plot functions of probability pssm

Setup

# Load data for testing
from katlas.data import *
from katlas.utils import *
from katlas.pssm import *

df=Data.cddm()
pssm_df = recover_pssm(df.loc['AKT1'])

Heatmap

It will convert s,t,y to pS,pT,pY in the heatmap.

---

plot_heatmap

def plot_heatmap(
    heatmap_df, ax:NoneType=None, position_label:bool=True, figsize:tuple=(5, 6), include_zero:bool=True,
    scale_pos_neg:bool=False, colorbar_title:str='Prob.', vmin:NoneType=None, vmax:NoneType=None, cmap:NoneType=None,
    center:NoneType=None, robust:bool=False, annot:NoneType=None, fmt:str='.2g', annot_kws:NoneType=None,
    linewidths:int=0, linecolor:str='white', cbar:bool=True, cbar_kws:NoneType=None, cbar_ax:NoneType=None,
    square:bool=False, xticklabels:str='auto', yticklabels:str='auto', mask:NoneType=None
):

Plot a heatmap of pssm.

This function visualizes a PSSM or log-odds matrix as a heatmap with diverging color scales centered at 0.

Color scale behavior:

• By default (scale_pos_neg=False), the colormap is centered at 0, but the full data range determines the color intensity:

  \[ \text{color range} = [\min(\text{data}), \max(\text{data})], \quad \text{with center at } 0 \]

  This is useful when you want to emphasize whether values are above or below zero, but without enforcing symmetry.

• If scale_pos_neg=True, the function uses a balanced diverging scale via TwoSlopeNorm, such that:

  \[ \text{min color} = \min(\text{data}), \quad \text{center} = 0, \quad \text{max color} = \max(\text{data}) \]

  The positive and negative ranges are scaled separately, ensuring that both ends of the heatmap have equal visual weight — especially helpful for symmetric data like log-odds matrices.

Additional visual features: - The center position (\(i = 0\)) can be masked out using include_zero=False.

plot_heatmap(pssm_df-0.3,scale_pos_neg=False,figsize=(15, 6));

plot_heatmap(pssm_df-0.3,scale_pos_neg=True,figsize=(15, 6));

plt.close('all')

Two heatmaps comparison

---

plot_two_heatmaps

def plot_two_heatmaps(
    pssm1, pssm2, kinase_name:str='Kinase', title1:str='CDDM', title2:str='PSPA', figsize:tuple=(4, 4.5),
    cbar:bool=True, scale_01:bool=False, cbar_fontsize:int=10, kwargs:VAR_KEYWORD
):

Plot two side-by-side heatmaps with black rectangle borders, titles on top, shared kinase label below, and only left plot showing y-axis labels.

pssm_df2 = pssm_df.loc[:,-5:5]

plot_two_heatmaps(pssm_df2,pssm_df2,'AKT','AKT1','AKT2')

Logo motif

To distinguish sty from STY, sty (lowercase) in the input df are automatically converted to pS,pT,pY in logo motif.

---

plot_logo_raw

def plot_logo_raw(
    pssm_df, ax:NoneType=None, title:str='Motif', ytitle:str='Bits', figsize:tuple=(10, 2)
):

Plot logo motif using Logomaker.

plot_logo_raw(pssm_df)

We can find the center name is in lower case, so need to change them

---

change_center_name

def change_center_name(
    df
):

Transfer the middle s,t,y to S,T,Y for plot if s,t,y have values; otherwise keep the original.

Now instead of s,t,y, the center name becomes S, T and Y:

change_center_name(pssm_df)[0]

aa
P    0.000000
G    0.000000
A    0.000000
C    0.000000
S    0.744152
T    0.248538
V    0.000000
I    0.000000
L    0.000000
M    0.000000
F    0.000000
Y    0.007310
W    0.000000
H    0.000000
K    0.000000
R    0.000000
Q    0.000000
N    0.000000
D    0.000000
E    0.000000
s    0.000000
t    0.000000
y    0.000000
Name: 0, dtype: float64

---

get_pos_min_max

def get_pos_min_max(
    pssm_df
):

Get min and max value of sum of positive and negative values across each position.

---

scale_zero_position

def scale_zero_position(
    pssm_df
):

Scale position 0 so that: - Positive values match the max positive column sum of other positions - Negative values match the min (most negative) column sum of other positions

This function rescales position 0 in a log-odds PSSM so that its total positive and negative stack heights match those of the most extreme positions on either side.

This ensures the central position visually matches the dynamic range of surrounding positions in log-odds logo plots.

---

scale_pos_neg_values

def scale_pos_neg_values(
    pssm_df
):

Globally scale all positive values by max positive column sum, and negative values by min negative column sum (preserving sign).

---

convert_logo_df

def convert_logo_df(
    pssm_df, scale_zero:bool=True, scale_pos_neg:bool=False
):

Change center name from s,t,y to S, T, Y in a pssm and scaled zero position to the max of neigbors.

---

get_logo_IC

def get_logo_IC(
    pssm_df
):

For plotting purpose, calculate the scaled information content (bits) from a frequency matrix, using log2(3) for the middle position and log2(len(pssm_df)) for others.

To visualize the motif using Logomaker, the scaled PSSM is computed by weighting each amino acid’s frequency at position \(i\) by the position’s information content:

\[ \text{PSSM\_scaled}_i(x) = P_i(x) \cdot \mathrm{IC}_i \]

This results in a matrix where the total stack height at each position equals the information content, and each letter’s height is proportional to its contribution. This is the standard format used by Logomaker to generate sequence logos.

get_logo_IC(pssm_df)

Position	-20	-19	-18	-17	-16	-15	-14	-13	-12	-11	...	11	12	13	14	15	16	17	18	19	20
aa
P	0.019253	0.011378	0.016632	0.014723	0.027341	0.017469	0.018061	0.020054	0.018999	0.021432	...	0.022264	0.016570	0.017585	0.025658	0.024024	0.024794	0.016453	0.018620	0.011564	0.017775
G	0.020457	0.011642	0.018254	0.019222	0.033016	0.012927	0.018462	0.023318	0.019358	0.020458	...	0.023748	0.019803	0.018025	0.017396	0.018739	0.023716	0.020762	0.014741	0.014455	0.016590
A	0.023666	0.015347	0.015820	0.018404	0.025278	0.020613	0.022074	0.022852	0.021867	0.025329	...	0.023748	0.018186	0.018904	0.027833	0.024505	0.019943	0.017237	0.011637	0.016383	0.018170
C	0.003209	0.003704	0.004462	0.003272	0.006191	0.004891	0.008428	0.000933	0.002151	0.003897	...	0.003463	0.005658	0.002198	0.004349	0.003844	0.003234	0.001959	0.003879	0.003212	0.003160
S	0.016044	0.007144	0.014198	0.012678	0.019087	0.009782	0.008830	0.018188	0.012188	0.020945	...	0.013853	0.016570	0.011431	0.013916	0.015376	0.017248	0.012928	0.014741	0.014134	0.013430
T	0.009627	0.005292	0.007707	0.016359	0.008254	0.010830	0.008428	0.010726	0.007170	0.008281	...	0.009400	0.008083	0.007474	0.010437	0.012012	0.015092	0.014495	0.009698	0.005782	0.013035
V	0.015242	0.008467	0.015415	0.021267	0.020635	0.013276	0.014850	0.012592	0.011471	0.007793	...	0.014843	0.020611	0.013189	0.021744	0.018258	0.022099	0.012536	0.014353	0.013170	0.008690
I	0.015643	0.007144	0.013792	0.013496	0.011865	0.008734	0.010034	0.010260	0.011113	0.009742	...	0.009400	0.011720	0.011431	0.006958	0.011532	0.011858	0.011752	0.013189	0.009637	0.010665
L	0.016044	0.012171	0.021905	0.022494	0.022183	0.018517	0.020067	0.026116	0.016848	0.023867	...	0.022759	0.016166	0.025059	0.020875	0.030751	0.026950	0.021546	0.021335	0.015098	0.022910
M	0.003209	0.006086	0.007707	0.002863	0.005159	0.005590	0.002809	0.005130	0.003226	0.007306	...	0.003463	0.002425	0.006155	0.005654	0.004324	0.003773	0.005876	0.003491	0.004818	0.004345
F	0.004412	0.008202	0.005679	0.008589	0.008254	0.007686	0.007224	0.007462	0.008962	0.005845	...	0.010885	0.009295	0.008793	0.007393	0.009610	0.007546	0.003526	0.006594	0.005140	0.007505
Y	0.003209	0.001852	0.002840	0.003272	0.007222	0.002795	0.004013	0.003265	0.003943	0.004871	...	0.004453	0.003637	0.003517	0.003479	0.003844	0.003773	0.005484	0.004267	0.002570	0.005135
W	0.004813	0.002646	0.001217	0.002863	0.002064	0.002446	0.002408	0.004664	0.002151	0.002435	...	0.001484	0.002021	0.002198	0.005219	0.003363	0.000539	0.002742	0.002327	0.006425	0.002765
H	0.003610	0.003175	0.005679	0.006544	0.007738	0.004542	0.004816	0.006529	0.008245	0.005358	...	0.003958	0.005254	0.009672	0.006958	0.005285	0.007007	0.004701	0.003491	0.004176	0.006320
K	0.015643	0.009261	0.018660	0.010633	0.027857	0.016420	0.014449	0.017255	0.014339	0.025816	...	0.021769	0.015762	0.027257	0.024354	0.017778	0.033418	0.021546	0.020947	0.012849	0.013825
R	0.016446	0.011642	0.022311	0.024539	0.030953	0.014324	0.016857	0.021453	0.016490	0.018996	...	0.035127	0.018186	0.027257	0.016961	0.025946	0.030184	0.016453	0.013577	0.012207	0.013035
Q	0.012033	0.008996	0.015415	0.009815	0.008770	0.010132	0.017258	0.012592	0.010754	0.018022	...	0.029685	0.011720	0.010551	0.007828	0.012493	0.016709	0.010969	0.013965	0.007709	0.008295
N	0.008423	0.005557	0.012575	0.012269	0.013929	0.009782	0.008027	0.013991	0.010396	0.013151	...	0.009895	0.016974	0.007474	0.009133	0.006727	0.009702	0.008227	0.007370	0.008673	0.011060
D	0.017649	0.008996	0.010547	0.013496	0.019087	0.013276	0.016054	0.020986	0.013622	0.025329	...	0.016327	0.010508	0.018465	0.013916	0.021622	0.026950	0.011361	0.012025	0.009958	0.015405
E	0.020858	0.010848	0.021905	0.020858	0.025794	0.016770	0.028897	0.033578	0.013622	0.039941	...	0.024738	0.023036	0.021103	0.020440	0.026907	0.026411	0.018020	0.019395	0.014776	0.020145
s	0.008022	0.009790	0.011764	0.009407	0.010318	0.008036	0.009231	0.014924	0.008245	0.012177	...	0.012864	0.009295	0.012310	0.008698	0.008168	0.011858	0.009010	0.015904	0.010922	0.017380
t	0.004412	0.004234	0.004057	0.003681	0.003095	0.004192	0.005218	0.003731	0.004302	0.003897	...	0.004948	0.002425	0.007034	0.002609	0.005766	0.005390	0.003526	0.003491	0.001927	0.001975
y	0.003209	0.001852	0.000811	0.001636	0.000000	0.000699	0.002007	0.001865	0.001792	0.003897	...	0.004453	0.002021	0.001319	0.003044	0.001441	0.001617	0.003134	0.002327	0.001927	0.001185

23 rows × 41 columns

---

plot_logo

def plot_logo(
    pssm_df, title:str='Motif', scale_zero:bool=True, ax:NoneType=None, figsize:tuple=(10, 1)
):

Plot logo of information content given a frequency PSSM.

# plot_logo(pssm_df,scale_zero=False,figsize=(10,1))

Set scale_zero to default True can have better vision of the side amino acids

plot_logo(pssm_df,title='Motif',figsize=(10,1))

plt.close('all')

Multiple logos

As multiple figures:

---

plot_logos_idx

def plot_logos_idx(
    pssms_df, idxs:VAR_POSITIONAL, figsize:tuple=(14, 1)
):

Plot logos of a dataframe with flattened PSSMs with index ad IDs.

pssms=Data.cddm()

plot_logos_idx(pssms,'AKT1','AKT2')

In one figure:

---

plot_logos

def plot_logos(
    pssms_df, count_dict:NoneType=None, # used to display n in motif title
    prefix:str='Motif', figsize:tuple=(14, 1)
):

Plot all logos from a dataframe of flattened PSSMs as subplots in a single figure.

plot_logos(pssms.head(2),prefix=None)

plt.close('all')

Logo motif + Heatmap

---

plot_logo_heatmap

def plot_logo_heatmap(
    pssm_df, # column is position, index is aa
    title:str='Motif', figsize:tuple=(17, 10), include_zero:bool=False
):

Plot logo and heatmap vertically

plot_logo_heatmap(pssm_df,title='Kinase',figsize=(17,10))