pssm.pspa

PSPA data visualization

Overview

This module provides tools for visualizing Position-Specific Scoring Matrices (PSSMs) from Positional Scanning Peptide Array (PSPA) kinase substrate specificity data.

---

Data Preprocessing

preprocess_pspa(pssm) — Prepares raw PSSM data for logo visualization. It removes the duplicate ‘s’ row (since it duplicates ‘t’), renames phospho-residues to ‘pS/pT’ and ‘pY’, and applies log₂ transformation relative to the median (excluding position 0, which is handled separately via scaling).

pssm = preprocess_pspa(
    pssm=raw_pssm,  # DataFrame with amino acids as index, positions as columns
)

---

Visualization Functions

plot_logo_pspa(row, title, figsize) — Generates a sequence logo from a single row of PSPA data. The letter heights represent log₂(value/median), showing amino acid preferences at each position.

plot_logo_pspa(
    row=pspa.loc['GSK3B'],  # single kinase row from Data.get_pspa()
    title='GSK3B',          # plot title
    figsize=(5, 2),         # figure dimensions (width, height)
)

plot_logo_heatmap_pspa(row, title, figsize, include_zero) — Creates a combined visualization with the sequence logo on top and a heatmap below, allowing detailed inspection of specificity values at each position.

plot_logo_heatmap_pspa(
    row=pspa.loc['GSK3B'],  # single kinase row from Data.get_pspa()
    title='GSK3B',          # plot title
    figsize=(6, 10),        # figure dimensions (width, height)
    include_zero=False,     # whether to include position 0 in heatmap
)

---

Normalization Functions (old version)

raw2norm(df, PDHK) — Normalizes raw PSPA data following the method from Johnson et al. (Nature 2022). Column-normalizes by sum of 17 randomized amino acids (excluding S, T, C), scales cysteine by its median, and sets S/T values to position medians.

norm_df = raw2norm(
    df=kinase_df,  # DataFrame with position as index, amino acids as columns
    PDHK=False,    # True for PDHK1/PDHK4 (uses 16 amino acids, also excludes Y)
)

get_one_kinase(df, kinase, normalize, drop_s) — Extracts and reshapes data for a single kinase from the stacked PSPA DataFrame, optionally applying normalization.

kinase_data = get_one_kinase(
    df=data,           # stacked DataFrame (kinase as index, position+aa as columns)
    kinase='PDHK1',    # kinase name to extract
    normalize=False,   # apply Johnson et al. normalization
    drop_s=True,       # drop 's' row (duplicate of 't' in PSPA)
)

get_logo(df, kinase) — End-to-end function that takes raw stacked PSPA data and generates a sequence logo for a specific kinase, including proper S/T ratio calculation at position 0.

get_logo(
    df=raw_pspa_df,  # stacked DataFrame with kinase as index, substrates as columns
    kinase='AAK1',   # kinase name to visualize
)

Setup

Plot

---

preprocess_pspa

def preprocess_pspa(
    pssm
):

Drop row s as it’s a duplicate of t; rename t to pS/pT; calculate np.log2(pssm/pssm.median())

pspa=Data.get_pspa()

row = pspa.loc['GSK3B']
pssm = recover_pssm(row.dropna())
pssm = preprocess_pspa(pssm)
pssm

Position	-5	-4	-3	-2	-1	0	1	2	3	4
aa
P	0.128793	0.103768	0.327105	0.377614	0.200697	0.000000	0.857330	-0.156606	-0.022523	0.020985
G	0.267518	0.232140	0.709128	0.215152	0.186051	0.000000	-0.070893	-0.132404	-0.032647	-0.138561
A	0.198756	0.026650	0.861131	0.210679	0.040110	0.000000	-0.073537	0.007280	-0.146221	-0.092757
C	0.281692	0.223789	0.041703	0.297615	-0.036342	0.000000	0.018745	0.043140	0.002481	0.013151
S	0.000000	-0.019696	-0.001366	-0.016950	0.000000	5.167707	-0.001258	0.000000	-0.002485	-0.013272
T	0.000000	-0.019696	-0.001366	-0.016950	0.000000	4.454046	-0.001258	0.000000	-0.002485	-0.013272
V	0.052905	-0.029645	0.009523	0.047180	-0.161873	0.000000	0.067572	-0.151192	-0.154488	-0.213174
I	0.000000	-0.135793	-0.045760	0.094081	-0.205347	0.000000	0.065169	-0.034469	-0.089637	-0.222392
L	0.041050	-0.054824	-0.194011	-0.100647	-0.161873	0.000000	-0.013903	-0.019595	-0.097586	-0.129861
M	-0.049839	-0.090829	-0.012337	-0.100647	0.076869	0.000000	-0.001258	-0.027013	0.100694	-0.026668
F	0.071672	-0.127756	-0.144893	-0.145823	-0.247330	0.000000	-0.049908	0.064236	0.098376	-0.092757
Y	-0.007367	0.021841	-0.051406	-0.160236	-0.140617	0.000000	-0.192106	0.000000	0.079700	0.114257
W	0.108489	0.059871	0.023019	0.016753	-0.091356	0.000000	0.001257	0.007280	0.012363	0.148034
H	0.024288	-0.019696	-0.026169	-0.095098	0.000000	0.000000	0.016259	0.199623	0.161902	0.089637
K	-0.032199	0.144157	0.057528	0.237314	0.094903	0.000000	0.067572	0.016930	0.036774	0.082169
R	-0.042252	0.080831	0.598097	0.647946	0.049388	0.000000	0.050669	0.054898	0.084392	-0.129861
Q	-0.163205	-0.090829	-0.015093	-0.022206	0.083658	0.000000	0.040921	0.112292	-0.002485	-0.013272
N	-0.227709	0.019431	0.001364	-0.016950	0.014284	0.000000	-0.018991	0.156618	-0.007469	0.097067
D	-0.271386	-0.049753	-0.001366	-0.040753	-0.210875	0.000000	-0.238837	-0.234608	-0.116305	0.015767
E	-0.271386	-0.238787	-0.054238	-0.075844	-0.261600	0.000000	-0.253757	-0.137747	0.315457	1.379934
pS/pT	-0.112011	0.206940	0.355120	0.106174	-0.114481	0.000000	0.055519	-0.004874	1.836501	6.163857
pY	-0.106727	0.196308	0.158811	0.259140	0.699151	0.000000	0.222213	0.047855	1.191558	1.496894

---

plot_logo_pspa

def plot_logo_pspa(
    row, title:str='Motif', figsize:tuple=(5, 2)
):

plot_logo_pspa(pspa.loc['GSK3B'],title='GSK3B')

---

plot_logo_heatmap_pspa

def plot_logo_heatmap_pspa(
    row, # row of Data.get_pspa()
    title:str='Motif', figsize:tuple=(6, 10), include_zero:bool=False
):

Plot logo and heatmap vertically

plot_logo_heatmap_pspa(pspa.loc['GSK3B'],title='GSK3B')

Others (old version)

---

raw2norm

def raw2norm(
    df:DataFrame, # single kinase's df has position as index, and single amino acid as columns
    PDHK:bool=False, # whether this kinase belongs to PDHK family
):

Normalize single ST kinase data

This function implement the normalization method from Johnson et al. Nature: An atlas of substrate specificities for the human serine/threonine kinome

Specifically, > - matrices were column-normalized at all positions by the sum of the 17 randomized amino acids (excluding serine, threonine and cysteine), to yield PSSMs. >- PDHK1 and PDHK4 were normalized to the 16 randomized amino acids (excluding serine, threonine, cysteine and additionally tyrosine) >- The cysteine row was scaled by its median to be 1/17 (1/16 for PDHK1 and PDHK4). >- The serine and threonine values in each position were set to be the median of that position. >- The S0/T0 ratio was determined by summing the values of S and T rows in the matrix (SS and ST, respectively), accounting for the different S vs. T composition of the central (1:1) and peripheral (only S or only T) positions (Sctrl and Tctrl, respectively), and then normalizing to the higher value among the two (S0 and T0, respectively, Supplementary Note 1)

This function is usually implemented with the below function, with normalize being a bool argument.

---

get_one_kinase

def get_one_kinase(
    df:DataFrame, # stacked dataframe (paper's raw data)
    kinase:str, # a specific kinase
    normalize:bool=False, # normalize according to the paper; special for PDHK1/4
    drop_s:bool=True, # drop s as s is a duplicates of t in PSPA
):

Obtain a specific kinase data from stacked dataframe

Retreive a single kinase data from PSPA data that has an format of kinase as index and position+amino acid as column.

data = Data.get_pspa_st()

get_one_kinase(data,'PDHK1')

aa	A	C	D	E	F	G	H	I	K	L	...	P	Q	R	S	T	V	W	Y	t	y
position
-5	0.0594	0.0625	0.0589	0.0550	0.0775	0.0697	0.0687	0.0590	0.0515	0.0657	...	0.0451	0.0424	0.0594	0.0594	0.0594	0.0573	0.1001	0.0775	0.0583	0.0658
-4	0.0618	0.0621	0.0550	0.0511	0.0739	0.0715	0.0598	0.0601	0.0520	0.0614	...	0.0637	0.0552	0.0617	0.0608	0.0608	0.0519	0.0916	0.0739	0.0528	0.0752
-3	0.0608	0.0576	0.0499	0.0423	0.0803	0.0580	0.0674	0.0687	0.0481	0.0667	...	0.0570	0.0532	0.0532	0.0584	0.0584	0.0588	0.1113	0.0803	0.0416	0.0553
-2	0.0587	0.0655	0.0470	0.0437	0.0790	0.0890	0.0787	0.0533	0.0440	0.0637	...	0.0500	0.0543	0.0616	0.0565	0.0565	0.0519	0.1082	0.0790	0.0327	0.0557
-1	0.0782	0.1009	0.0989	0.0426	0.0650	0.0695	0.0782	0.0496	0.0409	0.0578	...	0.0540	0.0500	0.0469	0.0594	0.0594	0.0514	0.0756	0.0650	0.0358	0.0433
0	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	NaN	NaN	NaN	1.0000	0.4886	NaN	NaN	0.0000	0.4886	0.0000
1	0.0400	0.0562	0.0394	0.0355	0.0735	0.0400	0.0502	0.1288	0.0390	0.1439	...	0.0379	0.0455	0.0455	0.0455	0.0455	0.0797	0.0784	0.0735	0.0336	0.0452
2	0.0496	0.0783	0.0643	0.0555	0.0720	0.1067	0.0684	0.0480	0.0505	0.0555	...	0.0564	0.0653	0.0695	0.0601	0.0601	0.0508	0.0672	0.0720	0.0414	0.0594
3	0.0486	0.0609	0.0938	0.0684	0.1024	0.0676	0.0544	0.0583	0.0388	0.0552	...	0.0686	0.0502	0.0561	0.0588	0.0588	0.0593	0.0641	0.1024	0.0539	0.0431
4	0.0565	0.0749	0.0631	0.0535	0.0732	0.0655	0.0664	0.0625	0.0496	0.0552	...	0.0677	0.0553	0.0604	0.0626	0.0626	0.0579	0.0864	0.0732	0.0548	0.0575

10 rows × 22 columns

---

get_logo

def get_logo(
    df:DataFrame, # stacked Dataframe with kinase as index, substrates as columns
    kinase:str, # a specific kinase name in index
):

Given stacked df (index as kinase, columns as substrates), get a specific kinase’s logo

This function is to replicate the motif logo from Johnson et al. Nature: An atlas of substrate specificities for the human serine/threonine kinome. Given raw PSPA data, it can output a motif logo.

# load raw PSPA data
# df = pd.read_csv('https://github.com/sky1ove/katlas_raw/raw/refs/heads/main/nbs/raw/pspa_st_raw.csv').set_index('kinase')
# df.head()
# get_logo(df, 'AAK1')