Skip to main content
affpapa.org
RU EN
skill Claude Code CX Codex Cursor GitHub Copilot

deepspot-m — Claude Code skill: what it does, how to install

Reads a 224x224 H&E tile and returns predicted log1p-CPM expression for the genes you ask about. Useful when you want spatial transcriptomics from a slide you have already scanned.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓ 
npx skills add K-Dense-AI/scientific-agent-skills --skill deepspot-m -a claude-code
Open on GitHub

What deepspot-m does

The skill hands DeepSpot-M to your agent. One histology tile goes in, one predicted value per queried gene comes out, laid on the same tile grid. The gene panel travels with the weights and is exposed as model.gene_names, covering ~19k protein-coding symbols; anything outside it raises a KeyError that names the offender.

Loading goes through DeepSpotM.from_pretrained with a source argument (evo2, orthrus, prott5, scgpt, apertus), which decides which frozen gene embedding builds the projections. One source per run. image_processor turns the PIL image into a tensor, predict_genes(batch, genes) gives values in the order of your symbol list.

Prerequisites: uv pip install deepspotm==1.0.0, Python 3.10+ and a PyTorch build matching your CUDA runtime. The weights are gated, and approval is not authentication, since hf auth login is a separate step on the machine that downloads. Code sits under PolyForm Noncommercial, weights under CC-BY-NC-SA-4.0, and the access form restricts eligibility to academic or nonprofit researchers without commercial affiliations. That makes a funded collaboration a real blocker, so check it before you start. The optional histolab 0.7.0 tiler needs its own Python 3.10 or 3.11 environment with OpenSlide.

What it helps with

  • predict gene expression from an H&E tile without sequencing
  • tile a whole slide and keep the prediction grid aligned
  • compare output across different gene embedding sources
  • export the expression matrix to AnnData/H5AD
  • catch a wrong-sized crop before it reaches the model

Example prompts for the agent

  • Load DeepSpot-M and predict `EPCAM` and `CD3D` for `tile.png`.
  • Tile `slide.svs` at 224x224, run every tile through DeepSpot-M and write the matrix to H5AD.
  • Run the same tiles through `source="scgpt"` and `source="evo2"` and compare.

How to install deepspot-m

1Skills CLI (Vercel)

npx skills add K-Dense-AI/scientific-agent-skills --skill deepspot-m -a claude-code

2Manually: put the files in place

~/.claude/skills/deepspot-m/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/deepspot-m/

From the same set120

The whole set →
scientific-visualization K-Dense-AI Generates and audits publication-ready scientific figures in Matplotlib, Seaborn, and Plotly, focusing on truthful data encoding and accessibility. Claude Code CX Codex Cursor +1 2,400installs scientific-writing K-Dense-AI This skill drafts and audits scientific manuscripts and reports while tying claims to verified evidence. Use it for sections, references, declarations, and submission checks. Claude Code CX Codex Cursor +1 2,400installs statistical-analysis K-Dense-AI Runs hypothesis tests with assumption checking, effect sizes, and APA-formatted reports. Covers t-tests, ANOVA, correlation, regression, non-parametric and Bayesian methods. Claude Code CX Codex Cursor +1 2,300installs scientific-brainstorming K-Dense-AI Structures early-stage research ideation through independent idea generation, transparent scoring, and adversarial review. Useful for ranking hypotheses without premature consensus. Claude Code CX Codex Cursor +1 2,300installs literature-review K-Dense-AI This skill searches and synthesizes research literature, checks citations, and prepares Markdown or PDF reviews. Use it for reproducible searches and screening. Claude Code CX Codex Cursor +1 2,300installs scientific-critical-thinking K-Dense-AI Evaluates scientific claims, research methodology, and evidence quality. Applies GRADE and Cochrane Risk of Bias frameworks. Claude Code CX Codex Cursor +1 2,300installs scientific-schematics K-Dense-AI Creates PNG drafts of scientific diagrams and runs a separate quality review. Use it for neural network architectures, flowcharts, system diagrams, and biological pathways. Claude Code CX Codex Cursor +1 2,200installs scikit-learn K-Dense-AI A skill for Python machine learning tasks: classification, regression, clustering, preprocessing, and building ML pipelines with scikit-learn 1.9.1. Claude Code CX Codex Cursor +1 2,200installs

Similar

firecrawl-research-papers firecrawl Discovers and synthesizes academic literature using Firecrawl Research. Builds structured reviews from PubMed, bioRxiv, medRxiv, and arXiv papers. Claude Code CX Codex Cursor +1 33,800installs manimce-best-practices adithya-s-k A skill that steers the agent through Manim Community Edition: scenes, MathTex, 3D, camera work. Useful once a project starts with from manim import *. Claude Code CX Codex Cursor +1 3,900installs anysearch anysearch-ai This skill gives an agent web search, domain-specific queries, parallel batch search, and page extraction. Use it when a task needs current information or a claim needs checking. Claude Code CX Codex Cursor +1 38,500installs academic-paper Imbad0202 This skill runs a 12-agent academic pipeline across 11 modes. It guides projects from literature searches and outlines to citation audits and rebuttal reviews. Claude Code CX Codex Cursor +1 10,800installs nature-figure Yuan1z0825 A skill that builds, revises and audits scientific figures for manuscripts in Python or R. Useful when plots go to submission, not to a report. Claude Code CX Codex Cursor +1 18,700installs academic-paper-reviewer Imbad0202 This skill reviews an academic paper through five role-based perspectives. Use it for peer review, revision checks, or a focused methodology assessment. Claude Code CX Codex Cursor +1 9,800installs nature-polishing Yuan1z0825 A skill that polishes, translates and tightens existing academic prose. Facts, terminology and evidence boundaries stay intact. Claude Code CX Codex Cursor +1 17,400installs nature-citation Yuan1z skill, refactored into static/dynamic layers Finds Nature and CNS-family papers for manuscript claims, maps claims to sources, and prepares a reference-manager export. Use it when the journal scope matters. Claude Code CX Codex Cursor +1 15,600installs

Questions about deepspot-m

How do I install deepspot-m in Claude Code?

Run npx skills add K-Dense-AI/scientific-agent-skills --skill deepspot-m -a claude-code. The skill lands at ~/.claude/skills/deepspot-m/SKILL.md.

What does deepspot-m do?

It produces virtual spatial transcriptomics: a histology tile goes in, predicted log1p-CPM values come out for genes from the ~19k-symbol panel.

Can it run on CPU?

Yes, CPU is supported and CUDA is optional. The gated weights are still required, and the machine has to be authenticated with hf auth login after approval.

See also

Project sponsors

Companies that keep this analytics open
Gold sponsor
Nutra.Club

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor