Models for age estimation from semen using sperm-specific age-related CpG markers
Accurate age estimation from semen has the potential to significantly narrow down the pool of unidentified suspects in sexual assault investigations. In a study, we trained seven models using methylation data of 21 sperm-specific age-related CpG (AR-CpG) markers in 253 sperm DNA samples (ages 22 years to 67 years). Given that these methylation data are generated using two methylation SNaPshot assays, designated as panel I and panel II, we named three of these models as panel I model (n = 11), panel II model (n = 10), and panels I & II model (n = 21), respectively. Additionally, we also provided four other simplified models, named 3-CpG model, 5-CpG model, 9-CpG model, and 14-CpG model, respectively.
The following steps are outlined to demonstrate the functionality of these seven models. These models are to be used with methylation data generated using methylation SNaPshot assays.
Before starting, download files required to run the seven models:
# https://github.com/XiaoChao369/Xiao-et-al-2023-age-semen-models
options(stringsAsFactors = FALSE)
rm(list = ls())
For convenience, set the directory containing the data or saving the results:
dirModel <- "J:/Xiao-et-al-2023-age-semen-models" # model files
dirData <- paste(dirModel, "/example_data", sep = "") # genotyping tables
dirResult <- paste(dirModel, "/example_results", sep = "") # for saving results
The panel I model is a support vector machine with polynomial kernel (svmPoly) model, including 11 AR-CpG markers, namely cg01789162, cg11262154, cg19998819, cg27231587, cg18037145, cg19983027, cg27111970, cg03634854, cg03030301, cg04119405, and cg25715498. This model can be used to estimate individual age from semen DNA or better from sperm DNA.
Prepare data:
# There are several ways to prepare data used for age estimation. One way is
# to use a GeneMapper ID/IDX Software to export the genotyping table or
# the sizing table as a '.txt' file. Another way is to make a '.txt' file
# organized as shown below and contain at least the columns named
# "Sample.Name", "Panel", "Marker", "Allele.1", "Allele.2", "Height.1",
# and "Height.2".
# Sample.Name Panel Marker Allele.1 Allele.2 Height.1 Height.2 ...
# SN001 Sperm_AR-CpG_Panle_I cg01789162 G A 213 1524 ...
# SN001 Sperm_AR-CpG_Panle_I cg11262154 G A 2291 284 ...
# SN001 Sperm_AR-CpG_Panle_I cg19998819 G A 1406 271 ...
# SN001 Sperm_AR-CpG_Panle_I cg27231587 G A 541 1715 ...
# SN001 Sperm_AR-CpG_Panle_I cg18037145 G A 1765 754 ...
# SN001 Sperm_AR-CpG_Panle_I cg19983027 G A 131 1721 ...
# SN001 Sperm_AR-CpG_Panle_I cg27111970 G A 396 2367 ...
# SN001 Sperm_AR-CpG_Panle_I cg03634854 G A 801 1874 ...
# SN001 Sperm_AR-CpG_Panle_I cg03030301 G A 701 1146 ...
# SN001 Sperm_AR-CpG_Panle_I cg04119405 G A 1670 1476 ...
# SN001 Sperm_AR-CpG_Panle_I cg25715498 G A 3351 383 ...
Run the panel I model on your data using the "panel_I_model.R" script provided. The only thing you have to do is to copy the prepared data file into the 'dirData' folder.
setwd(dirModel)
source('panel_I_model.R')
The output 'estimated_age' is in years and is automatically printed in the R console but is also saved as a '.csv' file, which can be found in the 'dirResult' folder.
print(estimated_age)
# Sample.Name Estimated.Age
# 1 SN001 47.62784
The panel II model is a neural network (neuralnet) model, including 10 sperm-specific AR-CpG markers, namely cg06304190, cg06979108, cg12837463, cg12277678, cg13872326, cg20602007, cg25187042, cg21843517, cg04123357, and cg24812634. This model can be used to estimate individual age from semen DNA or better from sperm DNA.
Prepare data:
# There are several ways to prepare data used for age estimation. One way is
# to use a GeneMapper ID/IDX Software to export the genotyping table or
# the sizing table as a '.txt' file. Another way is to maker a '.txt' file
# organized as shown below and contain at least the columns named
# "Sample.Name", "Panel", "Marker", "Allele.1", "Allele.2", "Height.1",
# and "Height.2".
# Sample.Name Panel Marker Allele.1 Allele.2 Height.1 Height.2 ...
# SN001 Sperm_AR-CpG_Panel_II BC C T NA 1311 ...
# SN001 Sperm_AR-CpG_Panel_II cg06304190 C T 1006 783 ...
# SN001 Sperm_AR-CpG_Panel_II cg06979108 C T 1743 646 ...
# SN001 Sperm_AR-CpG_Panel_II cg12837463 C T 1546 1703 ...
# SN001 Sperm_AR-CpG_Panel_II cg20828122 C T 128 1619 ...
# SN001 Sperm_AR-CpG_Panel_II cg12277678 C T 1003 495 ...
# SN001 Sperm_AR-CpG_Panel_II cg13872326 C T 1178 398 ...
# SN001 Sperm_AR-CpG_Panel_II cg20602007 C T 1237 422 ...
# SN001 Sperm_AR-CpG_Panel_II cg25187042 C T 1606 590 ...
# SN001 Sperm_AR-CpG_Panel_II cg21843517 C T 1459 649 ...
# SN001 Sperm_AR-CpG_Panel_II cg04123357 C T 2452 442 ...
# SN001 Sperm_AR-CpG_Panel_II cg24812634 C T 112 2789 ...
Run the panel II model on your data using the "panel_II_model.R" script provided. The only thing you have to do is to copy the prepared data file into the 'dirData' folder.
setwd(dirModel)
source('panel_II_model.R')
The output 'estimated_age' is in years and is automatically printed in the R console but is also saved as a '.csv' file, which can be found in the 'dirResult' folder.
print(estimated_age)
# Sample.Name Estimated.Age
# 1 SN001 48.80101
The panels I & II model is a Bayesian regularized neural network (brnn) model, including 21 sperm-specific AR-CpG markers, namely cg01789162, cg11262154, cg19998819, cg27231587, cg18037145, cg19983027, cg27111970, cg03634854, cg03030301, cg04119405, cg25715498, cg06304190, cg06979108, cg12837463, cg12277678, cg13872326, cg20602007, cg25187042, cg21843517, cg04123357, and cg24812634. This model can be used to estimate individual age from semen DNA or better from sperm DNA.
Prepare data:
# There are several ways to prepare data used for age estimation. One way is
# to use a GeneMapper ID/IDX Software to export the genotyping table or
# the sizing table as a '.txt' file. Another way is to maker a '.txt' file
# organized as shown below and contain at least the columns named
# "Sample.Name", "Panel", "Marker", "Allele.1", "Allele.2", "Height.1",
# and "Height.2".
# Sample.Name Panel Marker Allele.1 Allele.2 Height.1 Height.2 ...
# SN001 Sperm_AR-CpG_Panel_I cg01789162 G A 213 1524 ...
# SN001 Sperm_AR-CpG_Panel_I cg11262154 G A 2291 284 ...
# SN001 Sperm_AR-CpG_Panel_I cg19998819 G A 1406 271 ...
# SN001 Sperm_AR-CpG_Panel_I cg27231587 G A 541 1715 ...
# SN001 Sperm_AR-CpG_Panel_I cg18037145 G A 1765 754 ...
# SN001 Sperm_AR-CpG_Panel_I cg19983027 G A 131 1721 ...
# SN001 Sperm_AR-CpG_Panel_I cg27111970 G A 396 2367 ...
# SN001 Sperm_AR-CpG_Panel_I cg03634854 G A 801 1874 ...
# SN001 Sperm_AR-CpG_Panel_I cg03030301 G A 701 1146 ...
# SN001 Sperm_AR-CpG_Panel_I cg04119405 G A 1670 1476 ...
# SN001 Sperm_AR-CpG_Panel_I cg25715498 G A 3351 383 ...
# SN001 Sperm_AR-CpG_Panel_II BC C T NA 1311 ...
# SN001 Sperm_AR-CpG_Panel_II cg06304190 C T 1006 783 ...
# SN001 Sperm_AR-CpG_Panel_II cg06979108 C T 1743 646 ...
# SN001 Sperm_AR-CpG_Panel_II cg12837463 C T 1546 1703 ...
# SN001 Sperm_AR-CpG_Panel_II cg20828122 C T 128 1619 ...
# SN001 Sperm_AR-CpG_Panel_II cg12277678 C T 1003 495 ...
# SN001 Sperm_AR-CpG_Panel_II cg13872326 C T 1178 398 ...
# SN001 Sperm_AR-CpG_Panel_II cg20602007 C T 1237 422 ...
# SN001 Sperm_AR-CpG_Panel_II cg25187042 C T 1606 590 ...
# SN001 Sperm_AR-CpG_Panel_II cg21843517 C T 1459 649 ...
# SN001 Sperm_AR-CpG_Panel_II cg04123357 C T 2452 442 ...
# SN001 Sperm_AR-CpG_Panel_II cg24812634 C T 112 2789 ...
Run the panels I & II model on your data using the "panels_I_&_II_model.R" script provided. The only thing you have to do is to copy the prepared data file into the 'dirData' folder.
setwd(dirModel)
source('panels_I_&_II_model.R')
The output 'estimated_age' is in years and is automatically printed in the R console but is also saved as a '.csv' file, which can be found in the 'dirResult' folder.
print(estimated_age)
# Sample.Name Estimated.Age
# 1 SN001 48.6893
The 3-CpG model is a support vector machine with polynomial kernel (svmPoly) model, including 3 AR-CpG markers, namely cg21843517, cg12837463, and cg19998819. This model can be used to estimate individual age from semen DNA or better from sperm DNA.
Prepare data:
# There are several ways to prepare data used for age estimation. One way is
# to use a GeneMapper ID/IDX Software to export the genotyping table or
# the sizing table as a '.txt' file. Another way is to maker a '.txt' file
# organized as shown below and contain at least the columns named
# "Sample.Name", "Panel", "Marker", "Allele.1", "Allele.2", "Height.1",
# and "Height.2".
# Sample.Name Panel Marker Allele.1 Allele.2 Height.1 Height.2 ...
# SN001 Sperm_AR-CpG_Panel_I cg01789162 G A 213 1524 ...
# SN001 Sperm_AR-CpG_Panel_I cg11262154 G A 2291 284 ...
# SN001 Sperm_AR-CpG_Panel_I cg19998819 G A 1406 271 ...
# SN001 Sperm_AR-CpG_Panel_I cg27231587 G A 541 1715 ...
# SN001 Sperm_AR-CpG_Panel_I cg18037145 G A 1765 754 ...
# SN001 Sperm_AR-CpG_Panel_I cg19983027 G A 131 1721 ...
# SN001 Sperm_AR-CpG_Panel_I cg27111970 G A 396 2367 ...
# SN001 Sperm_AR-CpG_Panel_I cg03634854 G A 801 1874 ...
# SN001 Sperm_AR-CpG_Panel_I cg03030301 G A 701 1146 ...
# SN001 Sperm_AR-CpG_Panel_I cg04119405 G A 1670 1476 ...
# SN001 Sperm_AR-CpG_Panel_I cg25715498 G A 3351 383 ...
# SN001 Sperm_AR-CpG_Panel_II BC C T NA 1311 ...
# SN001 Sperm_AR-CpG_Panel_II cg06304190 C T 1006 783 ...
# SN001 Sperm_AR-CpG_Panel_II cg06979108 C T 1743 646 ...
# SN001 Sperm_AR-CpG_Panel_II cg12837463 C T 1546 1703 ...
# SN001 Sperm_AR-CpG_Panel_II cg20828122 C T 128 1619 ...
# SN001 Sperm_AR-CpG_Panel_II cg12277678 C T 1003 495 ...
# SN001 Sperm_AR-CpG_Panel_II cg13872326 C T 1178 398 ...
# SN001 Sperm_AR-CpG_Panel_II cg20602007 C T 1237 422 ...
# SN001 Sperm_AR-CpG_Panel_II cg25187042 C T 1606 590 ...
# SN001 Sperm_AR-CpG_Panel_II cg21843517 C T 1459 649 ...
# SN001 Sperm_AR-CpG_Panel_II cg04123357 C T 2452 442 ...
# SN001 Sperm_AR-CpG_Panel_II cg24812634 C T 112 2789 ...
Run the 3-CpG model on your data using the "3_CpG_model.R" script provided. The only thing you have to do is to copy the prepared data file into the 'dirData' folder.
setwd(dirModel)
source('3_CpG_model.R')
The output 'estimated_age' is in years and is automatically printed in the R console but is also saved as a '.csv' file, which can be found in the 'dirResult' folder.
print(estimated_age)
# Sample.Name Estimated.Age
# 1 SN001 47.86754
The 5-CpG model is a support vector machine with polynomial kernel (svmPoly) model, including 5 AR-CpG markers, namely cg21843517, cg12837463, cg19998819, cg13872326, and cg03634854. This model can be used to estimate individual age from semen DNA or better from sperm DNA.
Prepare data:
# There are several ways to prepare data used for age estimation. One way is
# to use a GeneMapper ID/IDX Software to export the genotyping table or
# the sizing table as a '.txt' file. Another way is to maker a '.txt' file
# organized as shown below and contain at least the columns named
# "Sample.Name", "Panel", "Marker", "Allele.1", "Allele.2", "Height.1",
# and "Height.2".
# Sample.Name Panel Marker Allele.1 Allele.2 Height.1 Height.2 ...
# SN001 Sperm_AR-CpG_Panel_I cg01789162 G A 213 1524 ...
# SN001 Sperm_AR-CpG_Panel_I cg11262154 G A 2291 284 ...
# SN001 Sperm_AR-CpG_Panel_I cg19998819 G A 1406 271 ...
# SN001 Sperm_AR-CpG_Panel_I cg27231587 G A 541 1715 ...
# SN001 Sperm_AR-CpG_Panel_I cg18037145 G A 1765 754 ...
# SN001 Sperm_AR-CpG_Panel_I cg19983027 G A 131 1721 ...
# SN001 Sperm_AR-CpG_Panel_I cg27111970 G A 396 2367 ...
# SN001 Sperm_AR-CpG_Panel_I cg03634854 G A 801 1874 ...
# SN001 Sperm_AR-CpG_Panel_I cg03030301 G A 701 1146 ...
# SN001 Sperm_AR-CpG_Panel_I cg04119405 G A 1670 1476 ...
# SN001 Sperm_AR-CpG_Panel_I cg25715498 G A 3351 383 ...
# SN001 Sperm_AR-CpG_Panel_II BC C T NA 1311 ...
# SN001 Sperm_AR-CpG_Panel_II cg06304190 C T 1006 783 ...
# SN001 Sperm_AR-CpG_Panel_II cg06979108 C T 1743 646 ...
# SN001 Sperm_AR-CpG_Panel_II cg12837463 C T 1546 1703 ...
# SN001 Sperm_AR-CpG_Panel_II cg20828122 C T 128 1619 ...
# SN001 Sperm_AR-CpG_Panel_II cg12277678 C T 1003 495 ...
# SN001 Sperm_AR-CpG_Panel_II cg13872326 C T 1178 398 ...
# SN001 Sperm_AR-CpG_Panel_II cg20602007 C T 1237 422 ...
# SN001 Sperm_AR-CpG_Panel_II cg25187042 C T 1606 590 ...
# SN001 Sperm_AR-CpG_Panel_II cg21843517 C T 1459 649 ...
# SN001 Sperm_AR-CpG_Panel_II cg04123357 C T 2452 442 ...
# SN001 Sperm_AR-CpG_Panel_II cg24812634 C T 112 2789 ...
Run the 5-CpG model on your data using the "5_CpG_model.R" script provided. The only thing you have to do is to copy the prepared data file into the 'dirData' folder.
setwd(dirModel)
source('5_CpG_model.R')
The output 'estimated_age' is in years and is automatically printed in the R console but is also saved as a '.csv' file, which can be found in the 'dirResult' folder.
print(estimated_age)
# Sample.Name Estimated.Age
# 1 SN001 49.59422
The 9-CpG model is a support vector machine with polynomial kernel (svmPoly) model, including 9 AR-CpG markers, namely cg21843517, cg12837463, cg19998819, cg13872326, cg03634854, cg27231587, cg04119405, cg20602007, and cg01789162. This model can be used to estimate individual age from semen DNA or better from sperm DNA.
Prepare data:
# There are several ways to prepare data used for age estimation. One way is
# to use a GeneMapper ID/IDX Software to export the genotyping table or
# the sizing table as a '.txt' file. Another way is to maker a '.txt' file
# organized as shown below and contain at least the columns named
# "Sample.Name", "Panel", "Marker", "Allele.1", "Allele.2", "Height.1",
# and "Height.2".
# Sample.Name Panel Marker Allele.1 Allele.2 Height.1 Height.2 ...
# SN001 Sperm_AR-CpG_Panel_I cg01789162 G A 213 1524 ...
# SN001 Sperm_AR-CpG_Panel_I cg11262154 G A 2291 284 ...
# SN001 Sperm_AR-CpG_Panel_I cg19998819 G A 1406 271 ...
# SN001 Sperm_AR-CpG_Panel_I cg27231587 G A 541 1715 ...
# SN001 Sperm_AR-CpG_Panel_I cg18037145 G A 1765 754 ...
# SN001 Sperm_AR-CpG_Panel_I cg19983027 G A 131 1721 ...
# SN001 Sperm_AR-CpG_Panel_I cg27111970 G A 396 2367 ...
# SN001 Sperm_AR-CpG_Panel_I cg03634854 G A 801 1874 ...
# SN001 Sperm_AR-CpG_Panel_I cg03030301 G A 701 1146 ...
# SN001 Sperm_AR-CpG_Panel_I cg04119405 G A 1670 1476 ...
# SN001 Sperm_AR-CpG_Panel_I cg25715498 G A 3351 383 ...
# SN001 Sperm_AR-CpG_Panel_II BC C T NA 1311 ...
# SN001 Sperm_AR-CpG_Panel_II cg06304190 C T 1006 783 ...
# SN001 Sperm_AR-CpG_Panel_II cg06979108 C T 1743 646 ...
# SN001 Sperm_AR-CpG_Panel_II cg12837463 C T 1546 1703 ...
# SN001 Sperm_AR-CpG_Panel_II cg20828122 C T 128 1619 ...
# SN001 Sperm_AR-CpG_Panel_II cg12277678 C T 1003 495 ...
# SN001 Sperm_AR-CpG_Panel_II cg13872326 C T 1178 398 ...
# SN001 Sperm_AR-CpG_Panel_II cg20602007 C T 1237 422 ...
# SN001 Sperm_AR-CpG_Panel_II cg25187042 C T 1606 590 ...
# SN001 Sperm_AR-CpG_Panel_II cg21843517 C T 1459 649 ...
# SN001 Sperm_AR-CpG_Panel_II cg04123357 C T 2452 442 ...
# SN001 Sperm_AR-CpG_Panel_II cg24812634 C T 112 2789 ...
Run the 9-CpG model on your data using the "9_CpG_model.R" script provided. The only thing you have to do is to copy the prepared data file into the 'dirData' folder.
setwd(dirModel)
source('9_CpG_model.R')
The output 'estimated_age' is in years and is automatically printed in the R console but is also saved as a '.csv' file, which can be found in the 'dirResult' folder.
print(estimated_age)
# Sample.Name Estimated.Age
# 1 SN001 49.76853
The 14-CpG model is a support vector machine with polynomial kernel (svmPoly) model, including 14 AR-CpG markers, namely cg21843517, cg12837463, cg19998819, cg13872326, cg03634854, cg27231587, cg04119405, cg20602007, cg01789162, cg06304190, cg25715498, cg03030301, cg04123357, and cg18037145. This model can be used to estimate individual age from semen DNA or better from sperm DNA.
Prepare data:
# There are several ways to prepare data used for age estimation. One way is
# to use a GeneMapper ID/IDX Software to export the genotyping table or
# the sizing table as a '.txt' file. Another way is to maker a '.txt' file
# organized as shown below and contain at least the columns named
# "Sample.Name", "Panel", "Marker", "Allele.1", "Allele.2", "Height.1",
# and "Height.2".
# Sample.Name Panel Marker Allele.1 Allele.2 Height.1 Height.2 ...
# SN001 Sperm_AR-CpG_Panel_I cg01789162 G A 213 1524 ...
# SN001 Sperm_AR-CpG_Panel_I cg11262154 G A 2291 284 ...
# SN001 Sperm_AR-CpG_Panel_I cg19998819 G A 1406 271 ...
# SN001 Sperm_AR-CpG_Panel_I cg27231587 G A 541 1715 ...
# SN001 Sperm_AR-CpG_Panel_I cg18037145 G A 1765 754 ...
# SN001 Sperm_AR-CpG_Panel_I cg19983027 G A 131 1721 ...
# SN001 Sperm_AR-CpG_Panel_I cg27111970 G A 396 2367 ...
# SN001 Sperm_AR-CpG_Panel_I cg03634854 G A 801 1874 ...
# SN001 Sperm_AR-CpG_Panel_I cg03030301 G A 701 1146 ...
# SN001 Sperm_AR-CpG_Panel_I cg04119405 G A 1670 1476 ...
# SN001 Sperm_AR-CpG_Panel_I cg25715498 G A 3351 383 ...
# SN001 Sperm_AR-CpG_Panel_II BC C T NA 1311 ...
# SN001 Sperm_AR-CpG_Panel_II cg06304190 C T 1006 783 ...
# SN001 Sperm_AR-CpG_Panel_II cg06979108 C T 1743 646 ...
# SN001 Sperm_AR-CpG_Panel_II cg12837463 C T 1546 1703 ...
# SN001 Sperm_AR-CpG_Panel_II cg20828122 C T 128 1619 ...
# SN001 Sperm_AR-CpG_Panel_II cg12277678 C T 1003 495 ...
# SN001 Sperm_AR-CpG_Panel_II cg13872326 C T 1178 398 ...
# SN001 Sperm_AR-CpG_Panel_II cg20602007 C T 1237 422 ...
# SN001 Sperm_AR-CpG_Panel_II cg25187042 C T 1606 590 ...
# SN001 Sperm_AR-CpG_Panel_II cg21843517 C T 1459 649 ...
# SN001 Sperm_AR-CpG_Panel_II cg04123357 C T 2452 442 ...
# SN001 Sperm_AR-CpG_Panel_II cg24812634 C T 112 2789 ...
Run the 14-CpG model on your data using the "14_CpG_model.R" script provided. The only thing you have to do is to copy the prepared data file into the 'dirData' folder.
setwd(dirModel)
source('14_CpG_model.R')
The output 'estimated_age' is in years and is automatically printed in the R console but is also saved as a '.csv' file, which can be found in the 'dirResult' folder.
print(estimated_age)
# Sample.Name Estimated.Age
# 1 SN001 48.82262
Chao Xiao [email protected]
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent