DAPE-PD: Diffusion adaptation of genomic foundation model enables tailored-generation of cis-regulatory DNA
Abstract
Genomic foundation models (GFMs) have transformed our ability to learn the evolutionary language encoded in genomes, yet their potential for programmable regulatory DNA engineering remains underexplored. Regulatory function depends on combinatorial interactions among distributed sequence elements, motivating generation with bidirectional context. Here, we introduce DAPE-PD (Diffusion Adaptation of Pretrained Evo for Promoter Design), which adapts the autoregressive Evo model into a bidirectional masked diffusion model without foundation-scale retraining. DAPE-PD combines pretrained genomic representations with iterative unmasking to generate regulatory DNA sequences. Following genus-specific fine-tuning on genome-derived intergenic sequences, DAPE-PD generated sequences that retained genus-associated compositional features despite low similarity to the fine-tuning datasets. Experimental validation identified functional de novo promoters, while in silico prioritization recovered active designs from small experimental candidate sets across three non-model gut commensals. This framework connects genomic pretraining with diffusion-based generation, extending GFMs from sequence understanding toward functional cis-regulatory DNA engineering in living cells.
Related
- Limitations of Genomic Foundation Models for Decoding Regulatory Mechanisms in ALS
- Where single-cell foundation models help: a benchmark of annotation, perturbation and regulatory inference against simple baselines
- Anthocyanins in woody ornamental plants: biosynthesis, regulatory networks, and trade-offs in stress adaptation
- Mendel, a foundation model of human genetic variation, prioritizes regulatory variants and improves gene expression prediction
- Distinct regulatory networks mediate plant adaptation to multifactorial abiotic stress combinations: rethinking of plant stress genetics
BibTeX
@misc{galindo2026dapepd,
title = {{DAPE-PD: Diffusion adaptation of genomic foundation model enables tailored-generation of cis-regulatory DNA}},
author = {M. Echavarria Galindo and S. Gong and Y. Zhan and S. Wang and L. Kong and Y. LAI},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.24.761598},
url = {https://www.biorxiv.org/content/10.64898/2026.10.24.761598v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - DAPE-PD: Diffusion adaptation of genomic foundation model enables tailored-generation of cis-regulatory DNA
AU - M. Echavarria Galindo
AU - S. Gong
AU - Y. Zhan
AU - S. Wang
AU - L. Kong
AU - Y. LAI
PY - 2026
DA - 2026/10/24
PB - bioRxiv
DO - 10.64898/2026.10.24.761598
UR - https://www.biorxiv.org/content/10.64898/2026.10.24.761598v1
AB - Genomic foundation models (GFMs) have transformed our ability to learn the evolutionary language encoded in genomes, yet their potential for programmable regulatory DNA engineering remains underexplored. Regulatory function depends on combinatorial interactions among distributed sequence elements, motivating generation with bidirectional context. Here, we introduce DAPE-PD (Diffusion Adaptation of Pretrained Evo for Promoter Design), which adapts the autoregressive Evo model into a bidirectional masked diffusion model without foundation-scale retraining. DAPE-PD combines pretrained genomic representations with iterative unmasking to generate regulatory DNA sequences. Following genus-specific fine-tuning on genome-derived intergenic sequences, DAPE-PD generated sequences that retained genus-associated compositional features despite low similarity to the fine-tuning datasets. Experimental validation identified functional de novo promoters, while in silico prioritization recovered active designs from small experimental candidate sets across three non-model gut commensals. This framework connects genomic pretraining with diffusion-based generation, extending GFMs from sequence understanding toward functional cis-regulatory DNA engineering in living cells.
ER -