GenomeQA: Benchmarking General Large Language Models for Genome Sequence Understanding

TL;DR

GenomeQA introduces a comprehensive benchmark to evaluate general-purpose large language models on raw genome sequence inference tasks, highlighting their strengths and limitations in biological sequence understanding.

Contribution

It provides a new controlled evaluation benchmark with diverse biological tasks and sequence lengths, enabling systematic study of LLMs in genomics.

Findings

Models outperform random baselines on sequence tasks.

Models exploit local signals like GC content and motifs.

Performance drops on complex, multi-step inference tasks.

Abstract

Large Language Models (LLMs) are increasingly adopted as conversational assistants in genomics, where they are mainly used to reason over biological knowledge, annotations, and analysis outputs through natural language interfaces. However, existing benchmarks either focus on specialized DNA models trained for sequence prediction or evaluate biological knowledge using text-only questions, leaving the behavior of general-purpose LLMs when directly exposed to raw genome sequences underexplored. We introduce GenomeQA, a benchmark designed to provide a controlled evaluation setting for general-purpose LLMs on sequence-based genome inference tasks. GenomeQA comprises 5,200 samples drawn from multiple biological databases, with sequence lengths ranging from 6 to 1,000 base pairs (bp), spanning six task families: Enhancer and Promoter Identification, Splice Site Identification, Taxonomic Classification, Histone Mark Prediction, Transcription Factor Binding Site Prediction, and TF Motif Prediction. Across six frontier LLMs, we find that models consistently outperform random baselines and can exploit local sequence signals such as GC content and short motifs, while performance degrades on tasks that require more indirect or multi-step inference over sequence patterns. GenomeQA establishes a diagnostic benchmark for studying and improving the use of general-purpose LLMs on raw genomic sequences.