[论文解读] Cross-Species Antimicrobial Resistance Prediction from Genomic Foundation Models

Huilin Tai|arXiv (Cornell University)|Mar 11, 2026

Antibiotic Use and Resistance被引用 0

一句话总结

本论文研究基因组基础模型在跨物种AMR预测中的应用，结果显示提取第10层嵌入并使用 MiniRocket 进行局部模式聚合可实现对 cassette 介导的耐药性转移的提升，且性能依赖于耐药机制。

ABSTRACT

Cross-species antimicrobial resistance (AMR) prediction is fundamentally an out-of-distribution (OOD) generalization problem: models trained on one set of bacterial taxa must transfer to phylogenetically distinct genomes that may rely on different resistance mechanisms. Across species, resistance arises from a heterogeneous mixture of localized, horizontally transferred gene cassettes and diffuse species-specific genomic backgrounds, making successful transfer inherently mechanism-dependent. Using a strict species holdout protocol, we first establish an interpretable k-mer baseline with Kover and show that strong within-species performance collapses under true cross-species evaluation. This motivates representation-level approaches that preserve transferable biological signals rather than amplify phylogenetic shortcuts. We investigate genomic foundation model embeddings derived from Evo-1-8k-base and introduce diagnostics for layer selection based on activation scale, isotropy, effective rank, and cross-seed stability under native bfloat16 inference. These analyses identify a stability boundary in deeper layers and reveal that embeddings extracted near this boundary provide more robust representations for downstream prediction. To preserve localized resistance signals, we treat per-window embeddings as an ordered multivariate signal and apply MiniRocket to summarize multi-scale local activation patterns instead of relying on global pooling. Our results show that aggregation strategy plays a central role in cross-species AMR prediction and that preserving local activation patterns substantially improves generalization when resistance mechanisms are localized.

研究动机与目标

• 在分布漂移下，探讨如何利用基因组基础模型实现跨物种 AMR 预测。
• 识别最能保留可迁移耐药信号的层和聚合策略。
• 理解耐药机制类型如何影响跨物种泛化。
• 提供一个可复现、抗泄漏的 AMR 模型评估框架。

提出的方法

• 利用 Evo 基因组基础模型嵌入，并基于诊断选择 Layer 10 的嵌入以提高稳定性和各向同性。
• 将每个窗口嵌入视为多变量信号，并应用 MiniRocket 进行局部模式聚合。
• 以严格的物种 holdout 分区基准对比 Kover 基线，评估真正的跨物种泛化。
• 筛选并整理一个包含五个分区和三个重复的多物种数据集，以实现鲁棒评估。
• 在 casssette 介导与染色体耐药机制之间比较聚合策略（MiniRocket vs Global Pooling）。

Figure 2.1: Category coverage among the top 30 targets, before (left) and after (right) filtering. The raw data shows extreme imbalance, with many antibiotics lacking sufficient species diversity. Post-filtering retains only targets that preserve diversity across all five partitions.

实验结果

研究问题

• RQ1 在严格的物种 holdout 下，基因组基础模型嵌入是否能迁移到未见物种？
• RQ2 哪一层与哪种聚合策略最能在不同物种间保留可迁移的耐药信号？
• RQ3 主导的耐药机制（cassette 介导 vs 染色体）如何影响跨物种泛化？
• RQ4 模型可解释性与邻居审计在跨物种 AMR 预测中的作用？

主要发现

• Kover 在同一物种内表现出强烈的性能，但在六种抗生素的物种 holdout 下跨物种性能显著下降。
• Layer 10 嵌入是最深且同时稳定的提取点，Layer 11 显示不稳定；层诊断界定了这一稳定性边界。
• 在 ampicillin 上，MiniRocket 搭配 k-NN 在 val_outside 划分通常获得最高的跨物种 MCC（0.753），优于使用 Global Pooling 的 k-NN。
• 在 ampicillin 的跨物种评估中，cassette 介导的耐药性带来更大的跨物种增益；染色体/弥散型耐药更难迁移。
• 在独立的 test_outside 集上，Global Pooling 与线性或基于树的分类器常常达到甚至超过 MiniRocket，具体取决于耐药机制。
• 当耐药信号局部化（cassette 介导）且在聚合时得以保留时，跨物种预测提升；同一物种的性能在方法之间保持相似。

Figure 2.2: Sample counts among the top 30 targets, before (left) and after (right) filtering. Many antibiotics in the raw data have insufficient samples for stable evaluation. Post-filter sets maintain adequate per-partition counts to yield reliable confidence intervals.