Skip to main content

Table 3 Results from the polygenic age-dependent model

From: A knowledge-based framework for the discovery of cancer-predisposing variants using large-scale sequencing breast cancer data

Variant

Approved name

Control MAF

Case MAF

Protein change

Mean beta elastic net

Negative beta percentage

MRPL24 - 1,156708335,C,T

Mitochondrial ribosomal protein L24

Novel

0.074%

W54*

−2.78

1.00

CST4 - 20,23667825,-,C

Cystatin S

0.0129%

0.300%

V81fs

−5.09

1.00

PARD6A - 16,67696278,C,T

Par-6 family cell polarity regulator alpha

0.0018%

0.078%

R256*

−1.86

1.00

TRIOBP - 22,38121788,-,C

TRIO and F-actin binding protein

0.0059%

0.471%

S1075fs

−3.64

1.00

ZNF85 - 19,21132125,C,T

Zinc finger protein 85

Novel

0.085%

R205*

−4.36

1.00

FOXP4 - 6,41553185,A,G

Forkhead box P4

0.0018%

0.091%

K147R

−8.04

1.00

PKHD1 - 6,51890490,A,C

Polycystic kidney and hepatic disease 1 (autosomal recessive)

Novel

0.075%

M1373R

−5.33

1.00

SURF1 - 9,136218808,A,T

Surfeit 1

Novel

0.081%

L179Q

−6.49

1.00

HIST2H2AB - 1,149859084,TT…GT,-

Histone cluster 2, H2ab

Novel

0.074%

T121fs

−3.59

0.97

STIM2 - 4,27004586,G,A

Stromal interaction molecule 2

Novel

0.081%

V281I

−1.65

0.97

CPA3 - 3,148597632,C,T

Carboxypeptidase A3 (mast cell)

Novel

0.074%

R178*

−5.47

0.94

TMCO3 - 13,114188422,-,G

Transmembrane and coiled-coil domains 3

0.0326%

0.742%

A469fs

−1.93

0.93

SERPINF2 - 17,1649022,CCTG,-

Serpin peptidase inhibitor, clade F

Novel

0.080%

A62fs

−1.74

0.84

PYGL - 14,51383751,G,A

Phosphorylase, glycogen, liver

0.0037%

0.149%

R276C

−0.08

0.71

FNIP2 - 4,159790466,C,A

Folliculin interacting protein 2

0.0016%

0.101%

S893*

−0.86

0.58

CPPED1 - 16,12758817,G,A

Calcineurin-like phosphoesterase domain containing 1

Novel

0.074%

R149*

−0.14

0.44

OR52B4 - 11,4388943,G,A

Olfactory receptor, family 52, subfamily B, member 4 (gene/pseudogene)

0.0018%

0.076%

R195*

4.81

0.09

SCN10A - 3,38755496,G,A

Sodium channel, voltage gated, type X alpha subunit

0.0037%

0.074%

R1155C

1.62

0.08

ZNF683 - 1,26694960,G,A

Zinc finger protein 683

Novel

0.089%

R35*

1.18

0.03

  1. A double-step machine learning algorithm selects variant based on a series of pathogenic prototypes and then further selects them using a permutation-based multi-model regression over age at onset. Variants in this set are negatively associated with age, and are divided in three layers: at the top, variants negatively associated in at least 80% of the models and with an average beta less than −1.5; in the middle, variants retained in at least 40% of the models with poor average beta; at the bottom, variants found negatively associated only in a few models
  2. *translation termination (stop) codon