Search papers, labs, and topics across Lattice.
The paper investigates gene expression prediction and finds that simply extending sequence length in current models degrades performance, even with careful algorithmic design. Instead, the authors emphasize the importance of integrating proximal multimodal epigenomic signals. They propose Prism, a framework that learns combinations of epigenomic features to represent chromatin states and uses backdoor adjustment to mitigate confounding effects, achieving state-of-the-art performance with short sequences.
Forget scaling sequence length: carefully integrating proximal epigenomic signals is the key to accurate gene expression prediction, outperforming long-sequence models.
Gene expression prediction, which predicts mRNA expression levels from DNA sequences, presents significant challenges. Previous works often focus on extending input sequence length to locate distal enhancers, which may influence target genes from hundreds of kilobases away. Our work first reveals that for current models, long sequence modeling can decrease performance. Even carefully designed algorithms only mitigate the performance degradation caused by long sequences. Instead, we find that proximal multimodal epigenomic signals near target genes prove more essential. Hence we focus on how to better integrate these signals, which has been overlooked. We find that different signal types serve distinct biological roles, with some directly marking active regulatory elements while others reflect background chromatin patterns that may introduce confounding effects. Simple concatenation may lead models to develop spurious associations with these background patterns. To address this challenge, we propose Prism, a framework that learns multiple combinations of high-dimensional epigenomic features to represent distinct background chromatin states and uses backdoor adjustment to mitigate confounding effects. Our experimental results demonstrate that proper modeling of multimodal epigenomic signals achieves state-of-the-art performance using only short sequences for gene expression prediction.
