A practical guide to building agentic systems for omics research. What tools do agents need? How do you handle the unique challenges of biological data? What does a reference architecture look like?
The central thesis of this series: Agentic Omics combines LLM reasoning with domain-specific biological AI models to create autonomous systems capable of end-to-end biological discovery.
Agentic AI is not just LLMs answering questions — it's AI systems that autonomously plan, reason, use tools, and execute multi-step scientific workflows. This post defines the paradigm and surveys its emergence in …
Single-cell technologies now measure multiple omics layers simultaneously in individual cells. AI models for single-cell multi-omics are enabling unprecedented resolution in understanding cellular heterogeneity, …
Phenomics bridges the molecular and the visible—using high-throughput imaging, clinical phenotypes, and electronic health records to connect genotype to phenotype. AI is central to this effort.
How deep learning and transformer models are transforming metagenomics, from assembly and taxonomic classification to predicting microbiome-disease associations and antibiotic resistance.
Metabolomics is the least 'language-model-ready' omics field, yet deep learning is transforming metabolite identification, pathway analysis, and biomarker discovery.
Protein AI is the most mature omics AI field. But structure prediction was just the beginning—the frontier is now protein design, function prediction, and interaction modeling.
Exploring the rise of single-cell foundation models like scGPT and Geneformer, spatial transcriptomics, and how transformer architectures are deciphering the complex language of gene expression.
DNA foundation models can now predict gene expression, identify regulatory elements, and classify variants from raw sequence alone. We review the major models—DNABERT-2, Nucleotide Transformer, Evo, and HyenaDNA—and what …
