# Modeling nascent transcription from chromatin landscape and structure with CLASTER

## TL;DR

CLASTER is a deep learning model that predicts nascent transcription levels using chromatin landscape and 3D structure data.

## Contribution

CLASTER introduces a novel deep neural network integrating chromatin data to predict transcription.

## Key findings

- CLASTER effectively translates chromatin data into kilobasepair-resolution transcription levels.
- The model reveals genomic organization as a key factor in machine learning approaches for transcription.
- CLASTER enables prediction of the impact of epigenetic perturbations in silico.

## Abstract

We present the Chromatin Landscape and Structure to Expression Regressor (CLASTER), an epigenetic-based deep neural network that can integrate different data modalities describing the chromatin landscape and its 3D structure. CLASTER effectively translates them into nascent transcription levels measured at a kilobasepair resolution. The model provides a platform to understand the epigenetic drivers and learned rules of nascent transcription, and to predict the impact of in silico epigenetic perturbations. We conclude that the predominant locality of current machine learning approaches emerges as a signature of genomic organization, having broad implications for future modeling approaches.

The online version contains supplementary material available at 10.1186/s13059-026-03992-5.