Software Development
DeepSpot M
DeepSpot‑M predicts log1p‑CPM expression for every protein‑coding gene across 224×224 tiles at roughly 20× magnification, generating transcriptome‑wide virtual spatial transcriptomics from H&E histology. The model queries a released gene panel by symbol and produces whole‑slide predictions through resolution‑aware tiling with histolab. Researchers and clinicians can therefore map gene expression patterns directly onto histological images without performing physical spatial transcriptomics assays.
- Publisher / source
- Scientific Agent Skills
- Upstream record
- K-Dense-AI/scientific-agent-skills
- Version
- Version 4
- Published
- Publication review
- Approved for publication
- Content fingerprint
a917434c523e136ddc1ce45b54acbef8c28398dd293314f72faa278431e67418