Title:	Creation of synthetic enhancers using a computational approach.
DNr:	NAISS 2025/22-1295
Project Type:	NAISS Small Compute
Principal Investigator:	Thomas Juan <thomas.juan@igp.uu.se>
Affiliation:	Uppsala universitet
Duration:	2025-09-23 – 2026-10-01
Classification:	10203
Keywords:

Abstract

Gene expression is modulated by cis-regulatory elements known as enhancers, which integrate transcription factor binding and chromatin context to control spatial and temporal transcription. Despite major advances in genome sequencing and epigenomics, predicting enhancer function and designing synthetic enhancers with precise activity remain significant challenges. This project aims to establish a computationally guided framework for the rational design of synthetic enhancers with defined regulatory properties. The approach integrates large-scale genomic and epigenomic datasets, including chromatin accessibility, histone modifications, and transcription factor binding profiles, with machine learning models. First, the regulatory context will be extracted using deep learning architectures trained on experimentally validated enhancers across multiple cell types. These models will identify critical sequence motifs and their combinatorial arrangements that drive enhancer activity. Second, computational algorithms will be developed to generate synthetic sequences that optimize these features while maintaining compatibility with endogenous chromatin landscapes. Predicted synthetic enhancers will be experimentally validated using massively parallel reporter assays and genome editing strategies. This iterative design, synthesis, and functional testing will refine the computational models and uncover fundamental principles of enhancer grammar and context dependency. By coupling in silico predictions with in vivo and in vitro assays, the project will provide unprecedented resolution on how motif composition, spacing, and epigenetic state influence enhancer strength and specificity. The creation of synthetic enhancers has broad scientific and practical implications. From a basic research perspective, it enables precise dissection of gene regulatory networks and the rules governing transcriptional control. From an applied standpoint, custom enhancers can be used to fine-tune gene expression in biotechnology, regenerative medicine, and gene therapy, allowing safe and predictable activation of therapeutic genes. Furthermore, the computational framework developed here will be adaptable to diverse organisms and cell types, facilitating the design of regulatory sequences for synthetic biology applications. This project combines computational biology, machine learning, and functional genomics to establish a new paradigm for enhancer engineering. The expected outcomes include: (1) predictive models of enhancer activity across specific cell types; and (2) validated synthetic enhancers with controllable strength and specificity. Collectively, these advances will deepen understanding of the non-coding genome and open avenues for precise, rational control of gene expression.