BreCol: Benchmarking Classical and Deep-Learning Methods for Microbiome-Based Cancer Detection
Abstract
DNA sequencing of the gut microbial community shows promise for cancer detection, but questions remain about the generalizability of results across studies. We propose BreCol, a benchmark of 2,040 16S rRNA gene sequencing runs across 26 studies spanning breast cancer, colorectal cancer, and healthy cohorts. Train-test splits are made within pre-2023 studies, while holdout evaluation uses studies from 2023 onward, reflecting temporal separation from training data. Classical models reach test/holdout AUCs of 0.77/0.60 for cancer diagnosis and 1.00/0.83 for cancer type prediction. We train the models on both cancer types simultaneously and find that colorectal cancer is often easier to detect than breast cancer. We also evaluate two deep learning models: HyenaDNA, a long-range sequence model that pools hidden states for classification, and SetBERT, a transformer that produces contextualized embeddings over sets of reads. Both deep learning models underperform the best classical methods on holdout data, though tuning training set size and the classification head yields modest gains. Our classical pipeline uses unsupervised clustering to derive features from tetramer frequencies, preserving within-run compositional signal and achieving near state-of-the-art performance without relying on taxonomic assignments. BreCol data and associated code are publicly available.
Related
- Benchmarking Microbiome Methods for Improved Metagenomic Sequencing of Lower Respiratory Tract Samples
- Detection of Frustration-related Operant Behavior in Rats via Machine Learning Methods
- OncoSpot enables benchmarking of cancer spatial transcriptomics methods across tumor evolutionary modes
- Benchmarking normalisation methods for differential binding analysis in CUT&RUN
- Benchmarking Recent Single-Cell Synthetic Data Generation Methods on the OneK1K Dataset
BibTeX
@misc{dick2026brecol,
title = {{BreCol: Benchmarking Classical and Deep-Learning Methods for Microbiome-Based Cancer Detection}},
author = {Jeffrey M. Dick},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.27207},
eprint = {2609.27207},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.27207},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - BreCol: Benchmarking Classical and Deep-Learning Methods for Microbiome-Based Cancer Detection
AU - Jeffrey M. Dick
PY - 2026
DA - 2026/09/23
PB - arXiv
DO - 10.48550/arxiv.2609.27207
UR - https://arxiv.org/abs/2609.27207
AB - DNA sequencing of the gut microbial community shows promise for cancer detection, but questions remain about the generalizability of results across studies. We propose BreCol, a benchmark of 2,040 16S rRNA gene sequencing runs across 26 studies spanning breast cancer, colorectal cancer, and healthy cohorts. Train-test splits are made within pre-2023 studies, while holdout evaluation uses studies from 2023 onward, reflecting temporal separation from training data. Classical models reach test/holdout AUCs of 0.77/0.60 for cancer diagnosis and 1.00/0.83 for cancer type prediction. We train the models on both cancer types simultaneously and find that colorectal cancer is often easier to detect than breast cancer. We also evaluate two deep learning models: HyenaDNA, a long-range sequence model that pools hidden states for classification, and SetBERT, a transformer that produces contextualized embeddings over sets of reads. Both deep learning models underperform the best classical methods on holdout data, though tuning training set size and the classification head yields modest gains. Our classical pipeline uses unsupervised clustering to derive features from tetramer frequencies, preserving within-run compositional signal and achieving near state-of-the-art performance without relying on taxonomic assignments. BreCol data and associated code are publicly available.
ER -