DNAmStress#

Index#

Instantiate model class
Define clock metadata
Download clock dependencies
Load features
Load weights into base model
Load reference values
Load preprocess and postprocess objects
Check all clock parameters
Basic test
Save torch model
Clear directory

Let’s first import some packages:

[1]:

import os
import inspect
import shutil
import json
import torch
import pandas as pd
import pyaging as pya

Instantiate model class#

[2]:

def print_entire_class(cls):
    source = inspect.getsource(cls)
    print(source)

print_entire_class(pya.models.DNAmStress)

class DNAmStress(LinearReferenceClock):
    pass

[3]:

model = pya.models.DNAmStress()

Define clock metadata#

[4]:

model.metadata["clock_name"] = "dnamstress"
model.metadata["data_type"] = "DNA methylation"  # Paper: Whole-blood DNA methylation was quantified for the epigenomic cohort.
model.metadata["species"] = "Homo sapiens"  # Paper: The NIAAA cohorts comprised human participants with AUD and healthy controls.
model.metadata["year"] = 2023
model.metadata["approved_by_author"] = "⌛"
model.metadata["citation"] = "Jung, J., et al. “Additive Effects of Stress and Alcohol Exposure on Accelerated Epigenetic Aging in Alcohol Use Disorder.” Biological Psychiatry 93(4): 331–341 (2023)."
model.metadata["doi"] = "https://doi.org/10.1016/j.biopsych.2022.06.036"
model.metadata["notes"] = "Whole-blood methylation score (MS stress) derived as a 211-CpG proxy for a composite of 13 stress-related measures."
model.metadata["research_only"] = None
model.metadata["tissue"] = ["whole blood"]  # Paper: Whole-blood DNAm was quantified in all 615 epigenomic-cohort participants.
model.metadata["predicts"] = ["stress exposure"]  # Paper: MS stress is a methylation-derived score tracking stress exposure.
model.metadata["training_target"] = ["stress exposure"]  # Paper: Three factor scores from 13 stress-associated measures were added to construct the composite score of stress.
model.metadata["unit"] = ["unitless"]  # Paper: MS stress was calculated as the weighted sum of selected CpG methylation values.
model.metadata["model_type"] = "bootstrap-stabilized elastic net regression"  # Paper: An elastic net model with 1,000 bootstraps selected CpGs present in more than half of bootstraps.
model.metadata["platform"] = ["Illumina EPIC"]  # Paper: Whole-blood DNAm was quantified using Infinium MethylationEPIC BeadChip arrays.
model.metadata["population"] = "adults"  # Paper: The epigenomic cohort consisted of 615 participants: 372 AUD cases and 243 healthy controls.
model.metadata["journal"] = "Biological Psychiatry"
model.metadata["last_author"] = "Falk W. Lohoff"
model.metadata["n_features"] = 211
model.metadata["citations"] = 27
model.metadata["citations_date"] = "2026-07-05"

Download clock dependencies#

[5]:

os.system(f"curl -sL -o DNAmStress_CpGs.rda https://raw.githubusercontent.com/HigginsChenLab/methylCIPHER/19b12296b0d7eb7055a97d068064df635f44ce3e/data/DNAmStress_CpGs.rda")
os.system(f"curl -sL -o DNAmStress_ref.rda https://raw.githubusercontent.com/HigginsChenLab/methylCIPHER/19b12296b0d7eb7055a97d068064df635f44ce3e/data/DNAmStress_ref.rda")

[5]:

[6]:

%%writefile download.r

library(jsonlite)
load("DNAmStress_CpGs.rda")
load("DNAmStress_ref.rda")
write_json(DNAmStress_CpGs, "coefficients.json", digits = 10)
write_json(DNAmStress_ref$intercept, "intercept.json", digits = 10)

Writing download.r

[7]:

os.system("Rscript download.r")

[7]:

Load features#

[8]:

coef_df = pd.DataFrame(json.load(open('coefficients.json')))
model.features = coef_df['CpG'].tolist()
iv = json.load(open('intercept.json'))
intercept_value = iv[0] if isinstance(iv, list) else iv

Load weights into base model#

[9]:

weights = torch.tensor(coef_df['coefficient'].tolist()).unsqueeze(0).float()
intercept = torch.tensor([intercept_value]).float()

[10]:

base_model = pya.models.LinearModel(input_dim=len(model.features))

base_model.linear.weight.data = weights.float()
base_model.linear.bias.data = intercept.float()

model.base_model = base_model

Load reference values#

[11]:

model.reference_values = None

Load preprocess and postprocess objects#

[12]:

model.preprocess_name = None
model.preprocess_dependencies = None

[13]:

model.postprocess_name = None
model.postprocess_dependencies = None

Check all clock parameters#

[14]:

pya.utils.print_model_details(model)


%==================================== Model Details ====================================%
Model Attributes:

training: True
metadata: {'approved_by_author': '⌛',
 'citation': 'Jung, Sun Jae, et al. "An epigenetic biomarker of stress and its '
             'association with cardiovascular disease." Biological Psychiatry '
             '93.4 (2023): 331-341.',
 'clock_name': 'dnamstress',
 'data_type': 'methylation',
 'doi': 'https://doi.org/10.1016/j.biopsych.2022.06.036',
 'notes': None,
 'research_only': None,
 'species': 'Homo sapiens',
 'version': None,
 'year': 2023}
reference_values: None
preprocess_name: None
preprocess_dependencies: None
postprocess_name: None
postprocess_dependencies: None
features: ['cg00188055', 'cg00418087', 'cg00452133', 'cg00599628', 'cg00684116', 'cg00728602', 'cg01047613', 'cg01085421', 'cg01245530', 'cg01607512', 'cg01631065', 'cg02107357', 'cg02386470', 'cg02606284', 'cg02606535', 'cg02607810', 'cg02652698', 'cg02873991', 'cg02978227', 'cg03104567', 'cg03120234', 'cg03221779', 'cg03579303', 'cg03581822', 'cg03626783', 'cg03642635', 'cg03693374', 'cg03723497', 'cg03744383', 'cg04048370']... [Total elements: 211]
base_model_features: None

%==================================== Model Details ====================================%
Model Structure:

base_model: LinearModel(
  (linear): Linear(in_features=211, out_features=1, bias=True)
)

%==================================== Model Details ====================================%
Model Parameters and Weights:

base_model.linear.weight: [1.7692569494247437, 5.452592849731445, 0.36093270778656006, 0.04217686131596565, -0.0875726118683815, -0.1617601066827774, -2.8009119033813477, -0.0006456300034187734, 2.6688179969787598, 1.0967479944229126, 2.873945951461792, 0.3544867932796478, -2.170973062515259, 2.942610025405884, 4.0788421630859375, -0.08575636893510818, 0.3709535002708435, 0.18792760372161865, -4.1162872314453125, -0.25866061449050903, 5.40001916885376, -2.2270829677581787, -0.7070323824882507, 1.7177339792251587, -0.6149824857711792, 0.26971670985221863, -1.411486029624939, 1.3798350095748901, -0.12567970156669617, 1.3220950365066528]... [Tensor of shape torch.Size([1, 211])]
base_model.linear.bias: tensor([-4.4941])

%==================================== Model Details ====================================%

Basic test#

[15]:

torch.manual_seed(42)
input = torch.randn(10, len(model.features), dtype=float)
model.eval()
model.to(float)
pred = model(input)
pred

[15]:

tensor([[-33.9411],
        [-62.7000],
        [ 93.3530],
        [ 14.1386],
        [-18.2623],
        [  4.9429],
        [ 94.5241],
        [ 63.0782],
        [  4.6987],
        [ 66.1088]], dtype=torch.float64, grad_fn=<AddmmBackward0>)

Save torch model#

[16]:

torch.save(model, f"../weights/{model.metadata['clock_name']}.pt")

Clear directory#

[17]:

# Function to remove a folder and all its contents
def remove_folder(path):
    try:
        shutil.rmtree(path)
        print(f"Deleted folder: {path}")
    except Exception as e:
        print(f"Error deleting folder {path}: {e}")

# Get a list of all files and folders in the current directory
all_items = os.listdir('.')

# Loop through the items
for item in all_items:
    # Check if it's a file and does not end with .ipynb
    if os.path.isfile(item) and not item.endswith('.ipynb'):
        os.remove(item)
        print(f"Deleted file: {item}")
    # Check if it's a folder
    elif os.path.isdir(item):
        remove_folder(item)

Deleted file: intercept.json
Deleted file: DNAmStress_CpGs.rda
Deleted file: download.r
Deleted file: coefficients.json
Deleted file: DNAmStress_ref.rda