Saving a MethylSeg Model

[1]:

%load_ext autoreload
%autoreload 2

[2]:

import sys

import numpy as np
import pandas as pd
import yaml
from pathlib import Path

import os


from methylseg import MethylSegPathway, MethylDataPrep, MethylStateAssignmentMethod, MethylationStates
from methylseg.helper_classes import DATA_DIR

[3]:

REFERENCE_DIR = DATA_DIR / Path("reference_files")

[4]:

tcga_sample_info, tcga_sample_info_removed = MethylDataPrep(
    meth_file=REFERENCE_DIR / "TCGA-BD-A3EP-01A_450k.tsv.gz",
    sample_id="TCGA-BD-A3EP-01A",
    resolution="450k",
    remove_low_coverage_like_cpgs=True,
).prepare()

[5]:

hm450k_model = MethylSegPathway(
    train_sample_info=tcga_sample_info,
    out_dir=Path("out") / "hm450k_model_outdir",
)

[6]:

hm450k_model.fit_pathway(force_optimize_rules=True)

Random search: 100%|██████████| 500/500 [04:43<00:00,  1.77it/s]

[7]:

hm450k_model.to_yaml(REFERENCE_DIR / "tcga_hm450k_model.yaml")

[8]:

test_model = MethylSegPathway.from_yaml(REFERENCE_DIR / "tcga_hm450k_model.yaml")

[9]:

wgbs_test_sample_info, wgbs_test_sample_info_removed = MethylDataPrep(
    meth_file=REFERENCE_DIR / "WGBS_colon-primary-tumor_1_wgbs.tsv.gz",
    sample_id="colon-primary-tumor_1",
    resolution="wgbs",
    min_coverage=10,
    remove_low_coverage_like_cpgs=True,
).prepare()

[10]:

wgbs_model = MethylSegPathway(
    train_sample_info=wgbs_test_sample_info,
    out_dir= Path("out") /  "wgbs_model_outdir",
    hmm_type="sticky"
)

[11]:

wgbs_model.fit_pathway(force_optimize_rules=True)

Random search: 100%|██████████| 500/500 [13:29<00:00,  1.62s/it]

[12]:

wgbs_model.to_yaml(REFERENCE_DIR / "wgbs_colon_model.yaml")

[13]:

test_model = MethylSegPathway.from_yaml(REFERENCE_DIR / "wgbs_colon_model.yaml")