Context-Aware Regularization with Markovian Integration for Attention-Based Nucleotide Analysis

TL;DR

This paper introduces CARMANIA, a novel self-supervised framework for nucleotide sequence analysis that integrates Markovian transition regularization into attention models, significantly improving long-range dependency capture and task performance.

Contribution

CARMANIA uniquely combines transition-matrix loss with attention mechanisms to explicitly model higher-order dependencies in genomic sequences, enhancing accuracy and efficiency.

Findings

Outperforms previous models by at least 7% on diverse tasks.

Matches or exceeds state-of-the-art on shorter sequences.

Achieves up to 34% MCC improvement in enhancer prediction.

Abstract

Transformers have revolutionized nucleotide sequence analysis, yet capturing long-range dependencies remains challenging. Recent studies show that autoregressive transformers often exhibit Markovian behavior by relying on fixed-length context windows for next-token prediction. However, standard self-attention mechanisms are computationally inefficient for long sequences due to their quadratic complexity and do not explicitly enforce global transition consistency. We introduce CARMANIA (Context-Aware Regularization with Markovian Integration for Attention-Based Nucleotide Analysis), a self-supervised pretraining framework that augments next-token (NT) prediction with a transition-matrix (TM) loss. The TM loss aligns predicted token transitions with empirically derived n-gram statistics from each input sequence, encouraging the model to capture higher-order dependencies beyond local context. This integration enables CARMANIA to learn organism-specific sequence structures that reflect both evolutionary constraints and functional organization. We evaluate CARMANIA across diverse genomic tasks, including regulatory element prediction, functional gene classification, taxonomic inference, antimicrobial resistance detection, and biosynthetic gene cluster classification. CARMANIA outperforms the previous best long-context model by at least 7 percent, matches state-of-the-art on shorter sequences (exceeding prior results on 20 out of 40 tasks while running approximately 2.5 times faster), and shows particularly strong improvements on enhancer and housekeeping gene classification tasks, including up to a 34 percent absolute gain in Matthews correlation coefficient (MCC) for enhancer prediction. The TM loss boosts accuracy in 33 of 40 tasks, especially where local motifs or regulatory patterns drive prediction.