MONTE-pretrained

MONTE pretrained models

This repository contains the pretrained models for MONTE, a Python package for tumor purity estimation and methylation correction using bulk DNA methylation data.

The models were trained using TCGA Illumina HumanMethylation450 data. This repository provides 21 cancer-specific models and one pan-cancer model.

Installation

Until MONTE is available on PyPI, install it from GitHub:

pip install "monte @ git+https://github.com/ylaboratory/MONTE.git"

Using uv:

uv add "monte @ git+https://github.com/ylaboratory/MONTE.git"

Usage

Load a cancer-specific model:

from monte import from_pretrained

model = from_pretrained("BRCA")

Or load the pan-cancer model:

model = from_pretrained("PAN-CANCER")

MONTE automatically downloads the requested model and stores it in the local Hugging Face cache.

For other MONTE usage, please refer to the MONTE repository and its tutorial.

Available models

ACC, BLCA, BRCA, CESC, COAD, GBM, HNSC, KICH, KIRC, KIRP, LGG,
LIHC, LUAD, LUSC, OV, PRAD, READ, SKCM, THCA, UCEC, UCS,
PAN-CANCER

The available models can also be listed in Python:

from monte import available_pretrained_models

print(available_pretrained_models())

License

The MONTE pretrained models are distributed under the BSD 3-Clause License. See LICENSE for details.

These models were trained using TCGA data. Users remain responsible for following the applicable TCGA and Genomic Data Commons data-use policies.

Citation

If you use MONTE or these pretrained models, please cite:

@article {
    Kim2026.01.22.701164,
    author = {Kim, Mirae and Lee, Wei-Hao and Yao, Vicky},
    title = {MONTE enables unified pan-cancer tumor purity estimation and methylation correction from bulk DNA methylation arrays},
    elocation-id = {2026.01.22.701164},
    year = {2026},
    doi = {10.64898/2026.01.22.701164},
    publisher = {Cold Spring Harbor Laboratory},
    URL = {https://www.biorxiv.org/content/early/2026/08/26/2026.01.22.701164},
    eprint = {https://www.biorxiv.org/content/early/2026/08/26/2026.01.22.701164.full.pdf},
    journal = {bioRxiv}
}