EPICGA#

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.EPICGA)

class EPICGA(LinearReferenceClock):
    def postprocess(self, x):
        """Model returns gestational age in days; convert to weeks."""
        return x / 7.0

[3]:

model = pya.models.EPICGA()

Define clock metadata#

[4]:

model.metadata["clock_name"] = "epicga"
model.metadata["data_type"] = "DNA methylation"  # Paper: An EPIC DNA-methylation predictor of gestational age.
model.metadata["species"] = "Homo sapiens"  # Paper: An EPIC DNA-methylation predictor of gestational age.
model.metadata["year"] = 2021
model.metadata["approved_by_author"] = "⌛"
model.metadata["citation"] = "Haftorn, K. L. et al. An EPIC predictor of gestational age and its application to newborns conceived by assisted reproductive technologies. Clinical Epigenetics 13, 82 (2021)."
model.metadata["doi"] = "https://doi.org/10.1186/s13148-021-01055-z"
model.metadata["notes"] = "LASSO predictor of ultrasound-estimated gestational age in days from umbilical cord-blood DNA methylation, trained in 755 non-ART START newborns."
model.metadata["research_only"] = None
model.metadata["tissue"] = ["cord blood"]  # Paper: Cord blood samples were taken immediately after birth.
model.metadata["predicts"] = ["gestational age"]  # Paper: An EPIC DNA-methylation predictor of gestational age.
model.metadata["training_target"] = ["gestational age"]  # Paper: Clinically estimated gestational age was regressed on CpGs.
model.metadata["unit"] = ["days"]  # Paper: Intercept 293.4557 and coefficients reproduce gestational age in days.
model.metadata["model_type"] = "LASSO regression"  # Paper: Lasso regression selected 176 CpGs.
model.metadata["platform"] = ["Illumina EPIC"]  # Paper: An EPIC DNA-methylation predictor of gestational age.
model.metadata["population"] = "newborns"  # Paper: The clock was trained on 755 non-ART newborns from START.
model.metadata["journal"] = "Clinical Epigenetics"
model.metadata["last_author"] = "Jon Bohlin"
model.metadata["n_features"] = 176
model.metadata["citations"] = 54
model.metadata["citations_date"] = "2026-07-05"

Download clock dependencies#

[5]:

supplementary_url = "https://static-content.springer.com/esm/art%3A10.1186%2Fs13148-021-01055-z/MediaObjects/13148_2021_1055_MOESM7_ESM.csv"
supplementary_file_name = "epic_ga_coefs.csv"
os.system(f"curl -sL -o {supplementary_file_name} {supplementary_url}")

[5]:

Load features#

[6]:

df = pd.read_csv('epic_ga_coefs.csv')
mask = df['cpgs'].astype(str).str.lower().isin(['intercept', '(intercept)'])
intercept_value = float(df.loc[mask, 's0'].iloc[0]) if mask.any() else 0.0
coef_df = df.loc[~mask].reset_index(drop=True)
model.features = coef_df['cpgs'].tolist()

Load weights into base model#

[7]:

weights = torch.tensor(coef_df['s0'].tolist()).unsqueeze(0).float()
intercept = torch.tensor([intercept_value]).float()

[8]:

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#

[9]:

model.reference_values = None

Load preprocess and postprocess objects#

[10]:

model.preprocess_name = None
model.preprocess_dependencies = None

[11]:

model.postprocess_name = 'days_to_weeks'
model.postprocess_dependencies = None

Check all clock parameters#

[12]:

pya.utils.print_model_details(model)


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

training: True
metadata: {'approved_by_author': '⌛',
 'citation': 'Haftorn, Kristine L., et al. "An EPIC predictor of gestational '
             'age and its application to newborns conceived by assisted '
             'reproductive technologies." Clinical Epigenetics 13.1 (2021): '
             '82.',
 'clock_name': 'epicga',
 'data_type': 'methylation',
 'doi': 'https://doi.org/10.1186/s13148-021-01055-z',
 'notes': None,
 'research_only': None,
 'species': 'Homo sapiens',
 'version': None,
 'year': 2021}
reference_values: None
preprocess_name: None
preprocess_dependencies: None
postprocess_name: 'days_to_weeks'
postprocess_dependencies: None
features: ['cg12701018', 'cg00078456', 'cg14958032', 'cg00735586', 'cg09035049', 'cg23915699', 'cg24741609', 'cg00996847', 'cg12079303', 'cg20301308', 'cg21141647', 'cg13189264', 'cg16246545', 'cg06902698', 'cg01281797', 'cg01833485', 'cg12434132', 'cg06046912', 'cg11147204', 'cg19672145', 'cg02968445', 'cg09116068', 'cg07749613', 'cg21155834', 'cg15433843', 'cg12685296', 'cg22871331', 'cg25241559', 'cg06900004', 'cg20582941']... [Total elements: 176]
base_model_features: None

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

base_model: LinearModel(
  (linear): Linear(in_features=176, out_features=1, bias=True)
)

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

base_model.linear.weight: [-0.09159819036722183, -2.3058528900146484, -0.4794044494628906, -11.715208053588867, -8.509488105773926, -0.5367807149887085, -0.32814541459083557, 1.0927908420562744, 1.626057744026184, -10.894654273986816, 1.7147382497787476, -8.59422492980957, 0.3499724566936493, 2.7312521934509277, 3.7085211277008057, 6.265093803405762, -3.1235597133636475, -0.5551669597625732, -4.767549991607666, 25.61239242553711, 1.6901012659072876, 1.5959044694900513, -32.38628387451172, -2.249823808670044, 11.97744369506836, 6.300745010375977, -0.7150956988334656, -0.45685043931007385, 3.390894889831543, -7.354003429412842]... [Tensor of shape torch.Size([1, 176])]
base_model.linear.bias: tensor([293.4557])

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

Basic test#

[13]:

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

[13]:

tensor([[60.5326],
        [42.4761],
        [51.9958],
        [58.1296],
        [27.9050],
        [42.0799],
        [33.6746],
        [32.6965],
        [63.3254],
        [42.7455]], dtype=torch.float64, grad_fn=<DivBackward0>)

Save torch model#

[14]:

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

Clear directory#

[15]:

# 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: epic_ga_coefs.csv