GBactPro: cross-species bacterial promoter prediction using element-resolved promoter definitions
- 1Department of Computer Science, National Chengchi University; Institute of Neuroscience, National Chengchi University
Abstract
Accurate bacterial promoter prediction supports genome annotation, but many models use fixed-length promoter windows anchored to transcription start sites (TSSs) rather than promoter-element boundaries. We developed the General Bacterial Promoter Predictor (GBactPro), a convolutional neural network-bidirectional long short-term memory (CNN-BiLSTM) framework using promoter sequences resolved by the Promoter Architecture Scanner. Using Escherichia coli-trained models evaluated across 43 bacterial species, we compared eight input designs and identified the contiguous -35/spacer/-10 core as optimal (mean AUROC = 0.9859). Species-specific models showed broad transfer in a 43 x 43 evaluation. Joint training on all 43 species achieved a mean AUROC of 0.9943 across their held-out test sets and outperformed four established predictors under matched-length benchmarks. In genome-wide E. coli scanning, GBactPro achieved AUROC and AUPRC values of 0.979 and 0.160 under stringent full-containment matching (0-bp tolerance), increasing to 0.997 and 0.323 under a 100-bp tolerance. False-positive predictions overlapped H-NS binding regions more frequently than true positives (24.4% versus 18.8%), and predictions also overlapped promoters derepressed in an H-NS-knockout dataset. Prediction precision varied across four TSS classes. These findings demonstrate that element-resolved promoter definitions improve cross-species prediction and support genome-wide identification of candidate bacterial promoters.
Related
- ProFinder: Candidate constitutive promoter prediction across the bacterial tree for genetic tool development in non-model species
- Octave: Scale-resolved Evaluation of Spatial Gene Expression Prediction from Histology
- Multi-vector retrieval enables residue-resolved prediction of protein partners by ColBERT-PPI
- Mutation-resolved ABL1 inhibitor resistance prediction under mutation-disjoint and temporal cross-study validation
- PathoBERT: A Hybrid Attention-Based Genomic Language Model for Read-Level Bacterial Pathogenicity Prediction
BibTeX
@misc{han2026gbactpro,
title = {{GBactPro: cross-species bacterial promoter prediction using element-resolved promoter definitions}},
author = {H.-Y. Han and Y.-C. V. Kao and S.-T. Kuo and H.-H. D. Chou and J.-M. Chang},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.18.759981},
url = {https://www.biorxiv.org/content/10.64898/2026.10.18.759981v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - GBactPro: cross-species bacterial promoter prediction using element-resolved promoter definitions
AU - H.-Y. Han
AU - Y.-C. V. Kao
AU - S.-T. Kuo
AU - H.-H. D. Chou
AU - J.-M. Chang
PY - 2026
DA - 2026/10/18
PB - bioRxiv
DO - 10.64898/2026.10.18.759981
UR - https://www.biorxiv.org/content/10.64898/2026.10.18.759981v1
AB - Accurate bacterial promoter prediction supports genome annotation, but many models use fixed-length promoter windows anchored to transcription start sites (TSSs) rather than promoter-element boundaries. We developed the General Bacterial Promoter Predictor (GBactPro), a convolutional neural network-bidirectional long short-term memory (CNN-BiLSTM) framework using promoter sequences resolved by the Promoter Architecture Scanner. Using Escherichia coli-trained models evaluated across 43 bacterial species, we compared eight input designs and identified the contiguous -35/spacer/-10 core as optimal (mean AUROC = 0.9859). Species-specific models showed broad transfer in a 43 x 43 evaluation. Joint training on all 43 species achieved a mean AUROC of 0.9943 across their held-out test sets and outperformed four established predictors under matched-length benchmarks. In genome-wide E. coli scanning, GBactPro achieved AUROC and AUPRC values of 0.979 and 0.160 under stringent full-containment matching (0-bp tolerance), increasing to 0.997 and 0.323 under a 100-bp tolerance. False-positive predictions overlapped H-NS binding regions more frequently than true positives (24.4% versus 18.8%), and predictions also overlapped promoters derepressed in an H-NS-knockout dataset. Prediction precision varied across four TSS classes. These findings demonstrate that element-resolved promoter definitions improve cross-species prediction and support genome-wide identification of candidate bacterial promoters.
ER -