Foundation Models¶
Pre-trained models for specialized omics domains.
Overview¶
Foundation models are large pre-trained models specialized for specific data types. Agentomics-ML can enable selected model families so the agent can use their catalog and local/downloaded weights during a run.
Available Types¶
| Type | Domain | Use Case |
|---|---|---|
dna |
Genomics | DNA sequences, variants |
rna |
Transcriptomics | RNA sequences, expression |
protein |
Proteomics | Protein sequences, structure |
molecule |
Chemistry | Small molecules, drugs |
Enabling Models¶
Enable foundation models before running:
This enables the selected family for the run and makes its catalog available to the agent. In Docker mode, the launcher pulls or builds an image tagged for that foundation-model type. In local mode, models are downloaded into the local workspace cache.
You can also use --foundation-models-type all to include every type.
Multiple Types¶
Enable a different type per run, or use all when the agent should see every configured type:
DNA Models¶
For genomic sequence data:
- Variant effect prediction
- Regulatory element detection
- Sequence classification
Example datasets: - Gene expression from DNA features - SNP effect prediction - Promoter classification
RNA Models¶
For transcriptomic data:
- RNA sequence analysis
- Secondary structure prediction
- Expression-based classification
Example datasets: - RNA-seq classification - Splice site prediction - RNA modification detection
Protein Models¶
For protein sequence data:
- Protein function prediction
- Structure-based classification
- Interaction prediction
Example datasets: - Protein family classification - Enzyme activity prediction - Binding site detection
Molecule Models¶
For small molecule/chemical data:
- Drug property prediction
- Molecular classification
- Activity prediction
Example datasets: - Drug-target interaction - Toxicity prediction - ADMET properties
How the Agent Uses Foundation Models¶
- Discovery - Agent queries available foundation models
- Selection - Agent chooses appropriate model for the data
- Embedding - Features extracted using the model
- Training - Embeddings used as input to ML model
Configuration¶
Foundation model configurations are in:
The YAML file lists each model family with:
- Model names and sources
- Domain type (dna, rna, protein, or molecule)
- Brief summaries
- Usage instructions for the agent
Without Foundation Models¶
If --foundation-models-type is omitted on a fresh run, no foundation models are made available to the agent. Forked runs inherit the source run's foundation-model type when the flag is omitted.
Storage Requirements¶
Foundation models can be large:
| Type | Approximate Size |
|---|---|
| DNA | 1-5 GB |
| RNA | 1-5 GB |
| Protein | 2-10 GB |
| Molecule | 0.5-2 GB |
Ensure sufficient disk space in the Docker volume or local environment.
GPU Acceleration¶
Foundation models benefit significantly from GPU:
- With GPU: Fast embedding generation
- CPU only: Much slower, but functional
Use --cpu-only if GPU unavailable, but expect longer run times for foundation model-based approaches.
Custom Foundation Models¶
To add custom foundation models:
- Add the family to
foundation_models/models.yaml - Add a companion usage guide under
foundation_models/ - Ensure the listed model names can be downloaded by
src/utils/download_foundation_models.py
Related¶
- Agent Architecture - How models are used
- GPU Settings - GPU configuration