HypoClock

Index

1. Instantiate model class

2. Define clock metadata

3. Download clock dependencies

4. Load features

5. Load weights into base model

6. Load reference values

7. Load preprocess and postprocess objects

8. Check all clock parameters

9. Basic test

10. Save torch model

11. Clear directory

Let’s first import some packages:

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

Instantiate model class

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

print_entire_class(pya.models.HypoClock)

class HypoClock(pyagingModel):
    def __init__(self):
        super().__init__()

    def preprocess(self, x):
        """
        Compute mean beta per sample, excluding missing (-1) values.
        """
        means = []
        for row in x:
            filtered_row = row[row != -1]
            if len(filtered_row) > 0:
                mean = torch.mean(filtered_row)
            else:
                mean = torch.tensor(float("nan"), device=x.device, dtype=x.dtype)
            means.append(mean)
        return torch.vstack(means)

    def postprocess(self, x):
        return 1 - x

model = pya.models.HypoClock()

Define clock metadata

model.metadata["clock_name"] = 'hypoclock'
model.metadata["data_type"] = 'methylation'
model.metadata["species"] = 'Homo sapiens'
model.metadata["year"] = 2018
model.metadata["approved_by_author"] = '⌛'
model.metadata["citation"] = "Zhou, Wanding, et al. \"DNA methylation loss in late-replicating domains is linked to mitotic cell division.\" Nature Genetics 50, no. 4 (2018): 591-602."
model.metadata["doi"] = "https://doi.org/10.1038/s41588-018-0073-4"
model.metadata["research_only"] = None
model.metadata["notes"] = "HypoClock score is 1 - mean beta across 678 solo-WCGW CpGs; reference values are -1 to ignore missing."

Download clock dependencies

Download directly with curl

supplementary_url = "https://raw.githubusercontent.com/aet21/EpiMitClocks/master/data/dataETOC3.rda"
supplementary_file_name = "dataETOC3.rda"
os.system(f"curl -L -o {supplementary_file_name} {supplementary_url}")

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 30851  100 30851    0     0   425k      0 --:--:-- --:--:-- --:--:--  430k

0

Load features

From R data

r_cmd = (
    "load('dataETOC3.rda'); "
    "df <- data.frame(cpg=dataETOC3.l[[4]]); "
    "write.csv(df, 'hypoclock_cpgs.csv', row.names=FALSE)"
)
os.system(f"Rscript -e \"{r_cmd}\"")

features_df = pd.read_csv('hypoclock_cpgs.csv')
model.features = features_df['cpg'].tolist()

assert len(model.features) == 678, f"Expected 678 CpGs, got {len(model.features)}"

Load weights into base model

weights = torch.tensor([1.0]).unsqueeze(0)
intercept = torch.tensor([0.0])

Linear model

base_model = pya.models.LinearModel(input_dim=1)

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

model.base_model = base_model

Load reference values

model.reference_values = [-1]*len(model.features)

Load preprocess and postprocess objects

model.preprocess_name = "mean"
model.preprocess_dependencies = None

model.postprocess_name = "one_minus"
model.postprocess_dependencies = None

Check all clock parameters

pya.utils.print_model_details(model)

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

training: True
metadata: {'approved_by_author': '⌛',
 'citation': 'Zhou, Wanding, et al. "DNA methylation loss in late-replicating '
             'domains is linked to mitotic cell division." Nature Genetics 50, '
             'no. 4 (2018): 591-602.',
 'clock_name': 'hypoclock',
 'data_type': 'methylation',
 'doi': 'https://doi.org/10.1038/s41588-018-0073-4',
 'notes': 'HypoClock score is 1 - mean beta across 678 solo-WCGW CpGs; '
          'reference values are -1 to ignore missing.',
 'research_only': None,
 'species': 'Homo sapiens',
 'version': None,
 'year': 2018}
reference_values: [-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1]... [Total elements: 678]
preprocess_name: 'mean'
preprocess_dependencies: None
postprocess_name: 'one_minus'
postprocess_dependencies: None
features: ['cg01689796', 'cg17753162', 'cg26150621', 'cg11684734', 'cg05812697', 'cg24052817', 'cg01827202', 'cg12277524', 'cg06160606', 'cg12758960', 'cg12270633', 'cg25939869', 'cg16823292', 'cg17041296', 'cg17325792', 'cg05347985', 'cg04789392', 'cg16320208', 'cg01779525', 'cg08009265', 'cg08244156', 'cg14711592', 'cg16577588', 'cg03114253', 'cg02245566', 'cg22601108', 'cg01068621', 'cg21291134', 'cg20491963', 'cg07725889']... [Total elements: 678]
base_model_features: None

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

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

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

base_model.linear.weight: tensor([[1.]])
base_model.linear.bias: tensor([0.])

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

Basic test

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

tensor([[0.9638],
        [0.9622],
        [1.0423],
        [1.0675],
        [1.0330],
        [0.9733],
        [0.9686],
        [1.0028],
        [0.9955],
        [0.9212]], dtype=torch.float64, grad_fn=<RsubBackward1>)

Save torch model

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

Clear directory

# 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: dataETOC3.rda
Deleted file: hypoclock_cpgs.csv