Research map / Large language models / Agents, code and math
SP-Mind: An Autonomous Reasoning Agent for Spatial Proteomics Analysis
YuCheng Yuan, Yuanfeng Ji, Zhongxiao Li, Ruijiang Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add expert-curated skill templates (ablation) | SP-Mind · SP-Bench average | execution accuracy (%): 62.4 -> 68.9 | Improved |
| add expert-curated skill templates (ablation) | SP-Mind · SP-Bench Advanced tier | execution accuracy (%): 44.4 -> 61.9 | Improved |
| add expert-curated skill templates (ablation) | SP-Mind · SP-Bench Basic tier | execution accuracy (%): 95.0 -> 95.8 | No clear effect |
| add expert-curated skill templates (ablation) | SP-Mind · CRC-CODEX Quantification | Pearson r: 0.945 -> 0.953 | No clear effect |
| add expert-curated skill templates (ablation) | SP-Mind · Cell Annotation PDAC IMC | CyteOnto GHK similarity: 0.762 -> 0.648 | Hurt |
| add expert-curated skill templates (ablation) | SP-Mind · Cell Annotation (4 datasets, average) | CyteOnto GHK similarity: 0.657 -> 0.681 | Improved |
| add expert-curated skill templates (ablation) | SP-Mind · Cell Annotation cHL 2 MIBI 5 | CyteOnto GHK similarity: 0.672 -> 0.765 | Improved |
| equip generalist agent with spatial proteomics toolkit | Biomni · SP-Bench average | execution accuracy (%): 19.3 -> 55.4 | Improved |
| equip generalist agent with spatial proteomics toolkit | ToolUniverse · SP-Bench average | execution accuracy (%): 15.0 -> 50.5 | Improved |
| use SP-Mind agent instead of Biomni-SP with same tools | Biomni-SP · SP-Bench Advanced tier | execution accuracy (%): 36.5 -> 61.9 | Improved |
| use SP-Mind instead of specialized ML baseline ASTIR on cell annotation | ASTIR · Cell Annotation (4 datasets, average) | CyteOnto GHK similarity: 0.440 -> 0.681 | Improved |
| longer multi-stage pipelines (Challenging tier) for all agents | generalist agents · SP-Bench Challenging tier | execution accuracy (%): 0.0 for AutoGen, Biomni, ToolUniverse; 33.3 for SP-Mind | Hurt |
Closest papers by meaning
- spora: A Unified Multimodal Dataset for Spatial Proteomics (NeurIPS 2026 · similarity 0.88)
- AutoScientists: Self-Organizing Agent Teams for Long-Running Scientific Experimentation (NeurIPS 2026 · similarity 0.83)
- SToFM: a Multi-scale Foundation Model for Spatial Transcriptomics (ICML 2025 · similarity 0.82)
- scPilot: Large Language Model Reasoning Toward Automated Single-Cell Analysis and Discovery (NeurIPS 2025 · similarity 0.82)
- STimage-1K4M: A histopathology image-gene expression dataset for spatial transcriptomics (NeurIPS 2024 · similarity 0.82)
- Diffusion Generative Modeling for Spatially Resolved Gene Expression Inference from Histology Images (ICLR 2025 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
