Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

DeepSpot-M Virtual Spatial Transcriptomics

A guide skill for running DeepSpot-M to predict transcriptome-wide spatial gene expression from H&E histology tiles.

Data & AnalyticsAdvanced★ 33,030⑂ 3,248AI score 8/10Last updated: Aug 9, 2026

What it does

  • Teaches Claude the correct way to run DeepSpot-M: feed a 224x224 RGB H&E tile at roughly 20x (~0.5 µm/pixel) and get per-gene expression in log1p-CPM.
  • Queries genes by HGNC symbol instead of a fixed panel, covering the released ~19k protein-coding gene set (including genes unseen in training).
  • Explains the five frozen gene-embedding sources (evo2, orthrus, prott5, scgpt, apertus) and how to choose between them.
  • Ships practical patterns: a tile-size guard, lazy import that turns ImportError into an actionable message, and a whole-slide batched inference loop that assembles a tiles-by-genes matrix into AnnData.

Who it's for

  • Computational biologists and grad students working in digital pathology or spatial transcriptomics.
  • Teams that want inferred expression across archival slide cohorts with no matching spatial assay.
  • ML engineers adding an expression channel to a morphology-only histology pipeline.
  • Note: noncommercial license only; requires gated Hugging Face weight access and PyTorch (CUDA GPU recommended).

Example uses

  1. "Predict EPCAM and CD3D for this tile" → generates the image_processor + predict_genes call with correct batching.
  2. "Tile this slide with histolab and build an expression map" → produces a loop that keeps tile coordinates, batches with torch.stack, and stores a tiles-by-genes matrix in AnnData.
  3. "Compare scgpt vs prott5 embedding sources" → suggests a script that runs the same tiles through two sources for sensitivity analysis.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/K-Dense-AI/scientific-agent-skills/HEAD/skills/deepspot-m/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills/deepspot-m folder from the GitHub repo K-Dense-AI/scientific-agent-skills into my ~/.claude/skills/deepspot-m/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/K-Dense-AI/scientific-agent-skills.git && mkdir -p ~/.claude/skills && cp -r scientific-agent-skills/skills/deepspot-m ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal and clone the skill repo: git clone https://github.com/K-Dense-AI/scientific-agent-skills.git
  2. Create the skills directory: mkdir -p ~/.claude/skills
  3. Copy just this skill: cp -r scientific-agent-skills/skills/deepspot-m ~/.claude/skills/
  4. Use Python 3.10–3.13 and install a PyTorch build matching your CUDA version first.
  5. Install the package: uv pip install deepspotm==1.0.0
  6. Request weight access at https://huggingface.co/ratschlab/DeepSpotM and wait for approval.
  7. Once granted, run huggingface-cli login on the machine that will download the weights (one time per machine).
  8. Restart Claude Code and try: "Use DeepSpot-M to predict gene expression for this H&E tile."
View source on GitHub ↗License: MIT