⚙
AI-generated deep summary
by claude@2026-07, 2026-07-04
· read from full text
ⓘ
The paper studies how to predict phage–bacterium interactions from DNA sequence data to reduce the labor required for experimental screening in phage therapy, using FoundedPBI, an ensemble deep learning framework built on genomic foundation models. The authors aggregate outputs from three DNA language models into a meta-embedding and then use a neural classifier, demonstrating that ensembling across prokaryotic and bacteriophage models captures complementary biological signals and improves F1-scores. They also adapt long-context aggregation strategies to process whole genomes up to 5 million base pairs, addressing the limitation that foundation models have much smaller context windows. A key limitation explicitly highlighted is that long-context challenges are largely unaddressed in prior genomic deep learning work, which they address, and the evaluation is primarily on the PredPHI benchmark and an internal CI4CB dataset; the paper does not explicitly discuss endometriosis or adenomyosis; it was included in the corpus via a keyword match in the upstream search index.
Abstract
The scalability of phage therapy as a viable alternative or complement to antibiotics is limited by the labor-intensive experimental screening required to identify compatible phage-bacterium pairs. To accelerate this discovery process, we propose FoundedPBI, an ensemble deep learning approach that leverages the emergent capabilities of genomic foundation models, large language models pre-trained on vast DNA corpuses to predict phage-bacterium interactions from DNA sequences alone. We employ an ensemble strategy that aggregates outputs from three state-of-the-art DNA language models into a unified meta-embedding, which is then processed by a neural classifier. Our approach makes two key contributions: (1) We demonstrate that performing ensemble learning across models trained on different genomic data—i.e., prokaryotic (Nucleotide Transformer v2, DNABERT-2) and bacteriophage (MegaDNA) genomes—captures partially-orthogonal biological signals, yielding 6% F1-score improvement over the best individual model. (2) We adapt long-context NLP aggregation strategies to handle whole bacterial and phage genomes (up to 5M base pairs) that exceed the foundation models’ context windows (12-96K bp) by a factor of 50–100, a critical challenge largely unaddressed in prior genomic deep learning work. On the PredPHI benchmark, FoundedPBI achieves a 76% F1-score outperforming the current state-of-the-art (PBIP) by 7%. On our internal dataset (CI4CB), we achieve 93% F1-score, improving our previous best methods by 4%. These results demonstrate that ensemble learning with proper long-context handling enables effective knowledge transfer of genomic foundation models to specialized prediction tasks.
Full text
1,789 characters
· extracted from
oa-doi-fallback
· click to expand
Abstract
The scalability of phage therapy as a viable alternative or complement to antibiotics is limited by the labor-intensive experimental screening required to identify compatible phage-bacterium pairs. To accelerate this discovery process, we propose FoundedPBI, an ensemble deep learning approach that leverages the emergent capabilities of genomic foundation models, large language models pre-trained on vast DNA corpuses to predict phage-bacterium interactions from DNA sequences alone. We employ an ensemble strategy that aggregates outputs from three state-of-the-art DNA language models into a unified meta-embedding, which is then processed by a neural classifier. Our approach makes two key contributions: (1) We demonstrate that performing ensemble learning across models trained on different genomic data—i.e., prokaryotic (Nucleotide Transformer v2, DNABERT-2) and bacteriophage (MegaDNA) genomes—captures partially-orthogonal biological signals, yielding 6% F1-score improvement over the best individual model. (2) We adapt long-context NLP aggregation strategies to handle whole bacterial and phage genomes (up to 5M base pairs) that exceed the foundation models’ context windows (12-96K bp) by a factor of 50–100, a critical challenge largely unaddressed in prior genomic deep learning work. On the PredPHI benchmark, FoundedPBI achieves a 76% F1-score outperforming the current state-of-the-art (PBIP) by 7%. On our internal dataset (CI4CB), we achieve 93% F1-score, improving our previous best methods by 4%. These results demonstrate that ensemble learning with proper long-context handling enables effective knowledge transfer of genomic foundation models to specialized prediction tasks.
Competing Interest Statement
The authors have declared no competing interest.
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.