Han#
Index#
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.Han)
class Han(pyagingModel):
def __init__(self):
super().__init__()
def preprocess(self, x):
return x
def postprocess(self, x):
"""
Applies an anti-logarithmic linear transformation to a PyTorch tensor.
"""
adult_age = 20
# Create a mask for negative and non-negative values
mask_negative = x < 0
mask_non_negative = ~mask_negative
# Initialize the result tensor
age_tensor = torch.empty_like(x)
# Exponential transformation for negative values
age_tensor[mask_negative] = (1 + adult_age) * torch.exp(x[mask_negative]) - 1
# Linear transformation for non-negative values
age_tensor[mask_non_negative] = (1 + adult_age) * x[mask_non_negative] + adult_age
return age_tensor
[3]:
model = pya.models.Han()
Define clock metadata#
[4]:
model.metadata["clock_name"] = 'han'
model.metadata["data_type"] = 'methylation'
model.metadata["species"] = 'Homo sapiens'
model.metadata["year"] = 2020
model.metadata["approved_by_author"] = '✅'
model.metadata["citation"] = "Han, Yang, et al. \"New targeted approaches for epigenetic age predictions.\" BMC biology 18 (2020): 1-15."
model.metadata["doi"] = "https://doi.org/10.1186/s12915-020-00807-2"
model.metadata["research_only"] = None
model.metadata["notes"] = None
Download clock dependencies#
[5]:
# from authors
cpg_sites = [
"(Intercept)",
'cg19283806',
'cg11807280',
'cg00329615',
'cg22454769',
'cg16867657',
'cg22796704',
'cg09809672',
'cg18618815',
'cg25533247',
'cg02286081',
'cg20222376',
'cg19344626',
'cg07082267',
'cg15845821',
'cg11741201',
'cg16054275',
'cg18933331',
'cg20249566',
'cg16604658',
'cg07583137',
'cg16008966',
'cg14556683',
'cg03746976',
'cg14314729',
'cg03431918',
'cg22156456',
'cg23078123',
'cg09748749',
'cg17457912',
'cg06492796',
'cg17593342',
'cg05308819',
'cg22512670',
'cg01820962',
'cg06639320',
'cg03224418',
'cg17436656',
'cg19500607',
'cg03735592',
'cg20669012',
'cg19761273',
'cg07080372',
'cg03638795',
'cg19722847',
'cg24711336',
'cg26935102',
'cg10221746',
'cg02085953',
'cg04604946',
'cg08558886',
'cg22361181',
'cg04208403',
'cg12623930',
'cg21572722',
'cg17885226',
'cg00748589',
'cg13033938',
'cg19784428',
'cg22016779',
'cg01974375',
'cg25256723',
'cg24724428',
'cg07547549',
'cg25410668',
'cg21296230'
]
coefficients = [
0.711184864,
-0.588354066,
-0.212038592,
0.014351188,
0.051285529,
2.152191741,
-0.689940565,
-0.643729974,
-0.772516118,
0.116662569,
-0.233409678,
0.002802259,
-0.062172432,
-0.224027294,
1.535209377,
0.344367661,
0.188826525,
-0.409150014,
-0.776065004,
0.500336643,
0.06125005,
-0.391624093,
0.100449175,
0.02000403,
0.266044453,
-0.259829677,
0.254063071,
-0.726178338,
-1.141947121,
-0.06322441,
-0.196926134,
0.85613244,
-0.887977059,
-0.334654336,
-0.854110638,
1.916122401,
0.92208575,
-0.070665617,
0.524707402,
0.319375235,
0.376055859,
0.033361038,
-1.458360975,
-0.267930475,
-0.590085273,
0.642506165,
0.470352872,
0.273581649,
-0.637989789,
-1.109388991,
-0.16886654,
0.662451226,
-0.091891613,
0.086290028,
-0.426089316,
0.32615363,
2.535639458,
-3.626802894,
0.097619541,
-0.427604263,
-0.41418774,
-0.27412342,
0.703772384,
-0.110027226,
0.283649813,
0.928585964
]
Load features#
[6]:
df = pd.DataFrame({
'feature': cpg_sites,
'coefficient': coefficients
})
model.features = df['feature'][1:].tolist()
Load weights into base model#
[7]:
weights = torch.tensor(df['coefficient'][1:].tolist()).unsqueeze(0)
intercept = torch.tensor([df['coefficient'][0]])
Linear model#
[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 = 'anti_log_linear'
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': 'Han, Yang, et al. "New targeted approaches for epigenetic age '
'predictions." BMC biology 18 (2020): 1-15.',
'clock_name': 'han',
'data_type': 'methylation',
'doi': 'https://doi.org/10.1186/s12915-020-00807-2',
'notes': None,
'research_only': None,
'species': 'Homo sapiens',
'version': None,
'year': 2020}
reference_values: None
preprocess_name: None
preprocess_dependencies: None
postprocess_name: 'anti_log_linear'
postprocess_dependencies: None
features: ['cg19283806', 'cg11807280', 'cg00329615', 'cg22454769', 'cg16867657', 'cg22796704', 'cg09809672', 'cg18618815', 'cg25533247', 'cg02286081', 'cg20222376', 'cg19344626', 'cg07082267', 'cg15845821', 'cg11741201', 'cg16054275', 'cg18933331', 'cg20249566', 'cg16604658', 'cg07583137', 'cg16008966', 'cg14556683', 'cg03746976', 'cg14314729', 'cg03431918', 'cg22156456', 'cg23078123', 'cg09748749', 'cg17457912', 'cg06492796']... [Total elements: 65]
base_model_features: None
%==================================== Model Details ====================================%
Model Structure:
base_model: LinearModel(
(linear): Linear(in_features=65, out_features=1, bias=True)
)
%==================================== Model Details ====================================%
Model Parameters and Weights:
base_model.linear.weight: [-0.5883540511131287, -0.21203859150409698, 0.014351188205182552, 0.051285527646541595, 2.152191638946533, -0.6899405717849731, -0.6437299847602844, -0.772516131401062, 0.11666256934404373, -0.23340967297554016, 0.0028022590558975935, -0.062172431498765945, -0.22402729094028473, 1.5352094173431396, 0.344367653131485, 0.1888265311717987, -0.40915000438690186, -0.7760649919509888, 0.5003366470336914, 0.061250049620866776, -0.3916240930557251, 0.10044917464256287, 0.0200040303170681, 0.2660444676876068, -0.25982967019081116, 0.25406306982040405, -0.7261783480644226, -1.1419471502304077, -0.0632244125008583, -0.19692613184452057]... [Tensor of shape torch.Size([1, 65])]
base_model.linear.bias: tensor([0.7112])
%==================================== 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([[ -1.0000],
[ 2.2557],
[310.9586],
[ 4.2231],
[ -0.7209],
[161.4707],
[ -0.7927],
[557.5405],
[ 0.8421],
[ -0.9992]], dtype=torch.float64, grad_fn=<IndexPutBackward0>)
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)