Checker and Scoring

from katlas.core import *

Checker

  • It will check if the middle position is s,t,y.
  • Check all sequences have same length
  • Convert non-s,t,y to upper case
  • Convert rare aa to _
df = Data.get_psp_human_site()
df.head()
gene protein uniprot site gene_site SITE_GRP_ID species site_seq LT_LIT MS_LIT MS_CST CST_CAT# Ambiguous_Site
0 YWHAB 14-3-3 beta P31946 T2 YWHAB_T2 15718712 human ______MtMDksELV NaN 3.0 1.0 None 0
1 YWHAB 14-3-3 beta P31946 S6 YWHAB_S6 15718709 human __MtMDksELVQkAk NaN 8.0 NaN None 0
2 YWHAB 14-3-3 beta P31946 Y21 YWHAB_Y21 3426383 human LAEQAERyDDMAAAM NaN NaN 4.0 None 0
3 YWHAB 14-3-3 beta P31946 T32 YWHAB_T32 23077803 human AAAMkAVtEQGHELs NaN NaN 1.0 None 0
4 YWHAB 14-3-3 beta P31946 S39 YWHAB_S39 27442700 human tEQGHELsNEERNLL NaN 4.0 NaN None 0 
try:
    check_seq_df(df, 'site_seq')
except Exception as e:
    print(f"Error in check_site: {e}")
Error in check_site: SSEVQFGhAGACANQ has h at position 7; need to have one of 's', 't', or 'y' in the center
df['site_seq'].str[7].value_counts()
site_seq
s    141851
t     58761
y     39367
h        14
k         4
r         3
g         3
p         2
n         1
f         1
l         1
a         1
i         1
d         1
Name: count, dtype: int64

Remove non-sty centered sequences:

df = df[df['site_seq'].str[7].isin(list('sty'))].copy()
df['site_seq'] = check_seq_df(df, 'site_seq')
df['site_seq']
0         ______MtMDKsELV
1         __MtMDKsELVQKAK
2         LAEQAERyDDMAAAM
3         AAAMKAVtEQGHELs
4         tEQGHELsNEERNLL
               ...       
240006    PsAKESAsQHITEEE
240007    GLPARPKsPLDPKKD
240008    LEQLLIKyPPEEVEs
240009    yPPEEVEsRRWQKIA
240010    NTAVEDAsDDESIPI
Name: site_seq, Length: 239979, dtype: object

Scoring

Params()
Available parameter sets:
['PSPA_st', 'PSPA_y', 'PSPA', 'CDDM', 'CDDM_upper']

Single sequence:

for param in ['PSPA', 'CDDM','CDDM_upper']:
    print(predict_kinase('PSVEPPLsQETFSDL',**Params(param)).head())
considering string: ['-5V', '-4E', '-3P', '-2P', '-1L', '0s', '1Q', '2E', '3T', '4F', '5S']
kinase
ATM       5.037
SMG1      4.385
DNAPK     3.818
ATR       3.507
FAM20C    3.170
dtype: float64
considering string: ['-7P', '-6S', '-5V', '-4E', '-3P', '-2P', '-1L', '0s', '1Q', '2E', '3T', '4F', '5S', '6D', '7L']
kinase
ATR      3.064
ATM      2.909
DNAPK    2.270
CK2A1    1.873
TSSK1    1.856
dtype: float64
considering string: ['-7P', '-6S', '-5V', '-4E', '-3P', '-2P', '-1L', '0S', '1Q', '2E', '3T', '4F', '5S', '6D', '7L']
kinase
ATR      3.229
ATM      3.038
DNAPK    2.479
CK2A1    2.006
CDK8     1.999
dtype: float64

Multiple sequences scoring

CDDM:

out_cddm = predict_kinase_df(df.head(500),'site_seq', **Params('CDDM'))
input dataframe has a length 500
Preprocessing
Finish preprocessing
Merging reference
Finish merging
CPU times: user 25.6 ms, sys: 4.73 ms, total: 30.3 ms
Wall time: 29.4 ms

PSPA:

out_cddm = predict_kinase_df(df.head(500),'site_seq', **Params('PSPA'))
input dataframe has a length 500
Preprocessing
Finish preprocessing
Merging reference
Finish merging
100%|██████████| 396/396 [00:00<00:00, 423.14it/s]
CPU times: user 973 ms, sys: 12.4 ms, total: 985 ms
Wall time: 998 ms

PSPA percentile score

Single sequence

st_pct = Data.get_pspa_st_pct()
y_pct = Data.get_pspa_tyr_pct()
a = get_pct('PSVEPPLyPETFSDL',**Params('PSPA_y'), pct_ref=y_pct)
a.sort_values('percentile',ascending=False)
considering string: ['-5V', '-4E', '-3P', '-2P', '-1L', '0Y', '1P', '2E', '3T', '4F', '5S']
log2(score) percentile
TEK 1.789 93.820916
PDHK4_TYR 0.851 87.204375
DDR2 0.204 80.929597
CSF1R 1.030 80.423787
PDHK3_TYR 0.477 79.384826
... ... ...
FLT1 -3.121 7.723855
TNNI3K_TYR -3.979 7.682843
PDGFRB -3.144 7.177033
PTK6 -3.540 4.743677
MUSK -5.376 3.773069

93 rows × 2 columns

get_pct('PSVEPPLsQETFSDL',**Params('PSPA_st'), pct_ref=st_pct)
considering string: ['-5V', '-4E', '-3P', '-2P', '-1L', '0S', '1Q', '2E', '3T', '4F']
log2(score) percentile
ATM 5.037 99.822351
SMG1 4.385 99.831819
DNAPK 3.818 99.205315
ATR 3.507 99.680344
FAM20C 3.170 95.370556
... ... ...
PKN1 -7.275 14.070436
P70S6K -7.295 4.089816
AKT3 -7.375 11.432995
PKCI -7.742 8.129511
NEK3 -8.254 4.637240

303 rows × 2 columns

Multiple sequences

score_df = predict_kinase_df(df.head(1000),'site_seq', **Params('PSPA_st'))
input dataframe has a length 1000
Preprocessing
Finish preprocessing
Merging reference
Finish merging
100%|██████████| 303/303 [00:00<00:00, 387.86it/s]
# get percentile reference
pct_ref = Data.get_pspa_st_pct()
pct = get_pct_df(score_df,pct_ref)
pct # the lower the better
100%|██████████| 303/303 [00:00<00:00, 1095.91it/s]
AAK1 ACVR2A ACVR2B AKT1 AKT2 AKT3 ALK2 ALK4 ALPHAK3 AMPKA1 ... VRK1 VRK2 WNK1 WNK3 WNK4 YANK2 YANK3 YSK1 YSK4 ZAK
0 73.188 87.606 84.729 79.877 79.533 81.850 89.768 91.201 98.426 54.554 ... 84.632 77.239 82.520 69.104 87.543 99.299 92.838 89.845 95.202 81.707
1 56.703 95.693 95.585 13.314 9.253 7.540 89.098 91.358 96.091 48.544 ... 30.292 25.050 17.132 47.388 20.742 80.747 87.061 8.621 53.530 45.247
2 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 ... 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000
3 73.105 82.606 85.710 62.441 67.236 58.350 83.195 88.883 72.216 62.304 ... 99.196 88.782 42.032 26.203 48.761 48.276 36.724 68.810 65.439 40.623
4 41.466 93.063 95.262 25.830 32.035 22.619 98.084 96.177 53.175 29.332 ... 31.234 6.206 19.819 38.389 18.550 79.177 45.058 10.560 42.568 26.996
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
995 78.612 48.108 49.316 70.027 68.095 73.951 54.940 60.196 67.759 27.998 ... 80.102 79.886 48.167 46.783 49.481 90.781 52.997 84.096 44.233 84.789
996 15.366 42.891 37.567 42.280 46.271 53.040 31.365 24.402 46.969 56.985 ... 65.541 68.537 28.641 52.885 39.271 75.061 76.763 52.028 81.469 74.303
997 37.572 21.732 22.480 83.025 76.807 79.376 14.507 46.188 3.412 88.018 ... 78.811 58.363 95.636 96.562 91.573 81.036 77.758 85.756 75.917 82.245
998 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 ... 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000
999 94.640 73.473 82.830 7.624 11.623 5.405 53.584 48.375 94.639 25.493 ... 87.950 66.768 8.615 0.937 14.440 14.019 3.233 73.357 35.516 42.516

1000 rows × 303 columns