Sign in to access this content
Model
DNABERT generates a dense representation of a genome sequence by identifying contextually similar sequences in the human genome.
Sign in to access this content
| Field | Response |
|---|---|
| Intended Application(s) & Domain(s): | Labeling of genomic sequences and comparing genomics |
| Model Type: | Genome Annotation |
| Intended Users: | This model is intended for developers or researchers who want to study the human genome sequence. |
| Output: | Sequence Embeddings |
| Describe how the model works: | Generate a dense representation of a genome sequence by identifying contextually similar sequences in the human genome. |
| Technical Limitations: | Model may not perform well on sequences that are highly divergent from the reference human genome assembly |
| Verified to have met prescribed NVIDIA standards: | Yes |
| Performance Metrics: | Validation Loss using cross entropy, F1 Harmonic Mean on the downstream task of splice site annotation. |
| Potential Known Risks: | The sequence embedding generated by model could lead to misleading inferences about a given sequence. |
| Licensing: | https://developer.download.nvidia.com/licenses/NVIDIA-BioNeMo-Framework-Evaluation-Software%20License(14Nov2023).pdf |