Single-Cell Genomics 2026 — Tools¶
Per-tool index covering the GRC Single-Cell Genomics 2026 (May 10–15, Les Diablerets, off-the-record) and the Wellcome Single Cell Biology 2026 (Jun 10–12, Hinxton, partially recorded). Both meetings together define the single-cell-methods cohort that AACR 2026 talks will cite for the next 12 months.
Status: Template stub + pre-meeting scaffolding. Wellcome speaker list dropped Apr 22, 2026 — ~15–20 tool pages can be pre-scaffolded from the speaker bios + their lab's most-recent preprint, then enriched post-meeting. GRC pages cannot be scaffolded in advance (talk list is closed); they get populated from preprint signals in the 4–8 weeks after May 15.
Per-tool entry template¶
Each tool gets its own page (tools/<package-slug>.md) following this structure. The single-cell-flavored extension of the standard EuroBioC / CSHL BDS template:
# <PackageName>
**One-line description** — what the package does in plain language.
- **Maintainer:** <name> (<institution>)
- **Modality:** scRNA / scATAC / spatial-transcriptomics / spatial-proteomics / CITE-seq / methylation / multimodal / lineage / atlas-platform
- **Language:** Python / R / both / standalone binary
- **Package:** [PyPI / Bioconductor / CRAN / GitHub-only]
- **Source:** [GitHub link]
- **Docs / vignette:** [link]
- **Preprint:** [bioRxiv DOI]
- **Publication:** [journal link, if any]
- **Status at SCG-2026:** new release / major update / methods talk / case study / atlas drop / preprint pre-announcement
- **Atlas footprint:** which large datasets it's been benchmarked on (HCA, Tabula Sapiens, CELLxGENE corpus, HuBMAP, HTAN, ENCODE-singlecell, MS-style atlases)
- **Foundation-model lineage:** if applicable — scGPT-line / Geneformer-line / scFoundation-line / scBERT-line / Universal Cell Embeddings / scPRINT / independent
## Talk at GRC Single-Cell Genomics 2026 (Les Diablerets, May 10–15)
- **Speaker:** ...
- **Session:** AI for Genomics / TF Regulation / Single-Cell Modeling / Applications in Medicine / Spatial & Proteomics / 3D Genomics / Epigenomics
- **Day:** ...
- **Slides / abstract:** Gordon Conferences are closed-door — no public record (extract from post-meeting preprints + attendee Bluesky/X with chair blessing)
- **Recording:** none (GRC policy)
## Talk at Wellcome Single Cell Biology 2026 (Hinxton, Jun 10–12)
- **Speaker:** ...
- **Session:** large scale / beyond transcriptomics / non-model / tech dev / computational / spatial
- **Day:** Wed 10 / Thu 11 / Fri 12
- **Slides:** [Zenodo or lab site]
- **Video:** [YouTube — Wellcome posts keynotes; selective for short talks]
- **Abstract page:** [Wellcome event page link]
## What it does
2–3 sentences: the problem it solves, key methodological idea, what it consumes/produces.
## Where it fits in the corpus
- AACR 2026 (single-cell + spatial topic): [link if mentioned]
- EuroBioC 2025: [link to SpatialData / DESpace / OSTA / spatialFDA / iSEE if related]
- CSHL BDS 2026 (spatial session): [link]
- RECOMB 2026 (ML for biology): [link]
- ISMB 2026 (MLCSB / RegSys / Sequence Analysis): [link]
- AGBT 2026 (platform-launch reference): [link if uses 10x Visium HD / Xenium / Curio Pixel / etc.]
## Notes
Free-form impressions, claimed benchmarks, head-to-head comparisons to alternatives (scGPT vs Geneformer vs scFoundation; Tangram vs cell2location vs CytoSPACE; etc.).
Tool index — pre-meeting scaffolding (Wellcome speakers only)¶
The 15 Wellcome speakers each anchor one likely tool/method/atlas. GRC speakers are unknown until on-site, so this table only pre-scaffolds the Wellcome half. The full ~30-tool table fills in post-Jun 12.
| Tool / method | Modality | Anchor speaker | Affiliation | Likely meeting slot | Cross-corpus links (anticipated) |
|---|---|---|---|---|---|
| scPRINT / CellFlow line | foundation model | Charlotte Bunne | EPFL | Wellcome — computational | AACR agentic-AI; RECOMB ML |
| ISS / In Situ Sequencing | spatial transcriptomics | Mats Nilsson | Stockholm | Wellcome — spatial | AACR spatial; AGBT platform |
| Monocle / sci-RNA-seq lineage | scRNA + lineage | Cole Trapnell | UW | Wellcome — organizer talk | AACR single-cell; EuroBioC |
| CellTypist / tissue atlases | atlas-scale annotation | Roser Vento-Tormo | Sanger | Wellcome — organizer + spatial | AACR atlas; HCA |
| CODEX / spatial-proteomics analysis | spatial proteomics | Sizun Jiang | Harvard | Wellcome — spatial | AACR spatial biomarkers |
| MIBI / multiplexed ion beam | spatial proteomics | Leeat Keren | Weizmann | Wellcome — organizer / large-scale | AACR spatial; Vallot Curie |
| Plasmodium atlas / MalariaCellAtlas | non-model scRNA | Mara Lawniczak | Sanger | Wellcome — non-model | none direct |
| DNA Typewriter / lineage barcoding | lineage tracing | James Gagnon | Utah | Wellcome — beyond transcriptomics | EMBO Lineage Tracing 2026 |
| Recording lineage (engineered) | lineage tracing | Theresa Loveless | Rice | Wellcome — tech dev | EMBO Lineage Tracing 2026 |
| novoSpaRc / spatial reconstruction theory | spatial-reconstruction algorithm | Mor Nitzan | Hebrew U | Wellcome — computational | EuroBioC SpatialData |
| Single-cell epigenomics (cancer) | scATAC / methylation in cancer | Céline Vallot | Curie | Wellcome — beyond transcriptomics | AACR epigenetics; HTAN |
| Oligodendrocyte single-cell atlas | scRNA / neuro | Gonçalo Castelo-Branco | Karolinska | Wellcome — large-scale | none direct |
| Lipid single-cell (D'Angelo) | beyond-transcriptomics | Giovanni D'Angelo | EPFL | Wellcome — beyond transcriptomics | none direct |
| Marlow comparative atlases | non-model / evo-devo | Heather Marlow | Chicago | Wellcome — non-model | none direct |
| Amit-line immune atlases / MARS-seq successors | scRNA / immune | Ido Amit | Weizmann | Wellcome — keynote | AACR immune-stromal; HCA |
Tool index — GRC (post-meeting, May 15+)¶
The GRC's seven sessions (AI for Genomics, TF Regulation, Single-Cell Modeling, Applications in Medicine, Spatial & Proteomics, 3D Genomics, Epigenomics) will each yield ~5 oral talks → ~30–35 tool candidates. Tool pages get drafted from:
- Preprints that drop May 15–Jul 15 with "presented at the 2026 Gordon Research Conference on Single-Cell Genomics" in the acknowledgements.
- Bluesky posts from attendees with the chair's permission (Pisco has historically blessed selective sharing).
- Selected speakers' lab Bluesky/X feeds announcing their published-by-then work.
Expected dominant lineages at GRC 2026 (based on the AI/ML subtitle and the field's mid-2026 trajectory):
- Foundation-model successors: scGPT-v2, Geneformer-v2, scFoundation-clinical, Universal Cell Embeddings updates, scBERT successors, perturbation-foundation models (Bunne's CellFlow class).
- Atlas-platform layer: CELLxGENE Discover updates, HCA v2.0, Tabula Sapiens v3, HuBMAP integration, HTAN cancer-atlas drops.
- Spatial inference: Visium HD analysis tooling (Voyager, SpatialData, semla, STUtility), Xenium-native pipelines, CosMx integration, MERSCOPE / MERFISH-Vizgen analysis.
- 3D genomics: scHi-C analysis (Nagano-line, Higashi successors), single-cell 3D contact-map inference, A/B compartment dynamics.
- Multimodal: CITE-seq / TEA-seq / Multiome integration, methylation+RNA, proteo-genomics.
Why this format¶
- Cross-vault links matter most at this meeting. Tools released here cycle through AACR sessions in the same year (talks reference "new methods like X") and AACR posters the next year (posters use X as a default). Single-Cell Genomics → AACR is the densest tool-to-clinical-application link in the corpus.
- Modality field is critical for slicing — single-cell is too broad a category for one-axis searching. scRNA-only methods vs spatial-only methods vs multimodal methods don't compete with each other.
- Atlas footprint captures the "is this tool already validated at scale?" dimension that distinguishes 2024-style methods (single dataset benchmarks) from 2026-style methods (CELLxGENE-corpus benchmarks expected).
- Foundation-model lineage captures the most important methods-evolution axis of 2026 — which transformer family is this descended from, and what does it cite as baseline?
- Two-meeting Talk fields are necessary because many tools will be presented at both GRC and Wellcome (slightly different versions/framings to slightly different audiences).
Open questions for the pilot¶
- GRC scaffolding from non-public sources — Pisco's discretion. Default policy: only include a GRC tool page if (a) a preprint exists by post-meeting drafting time, (b) the author has publicly acknowledged presenting at GRC, OR (c) we hear the talk content from an attendee with permission to share. No second-hand "I heard X presented Y" without confirmation.
- Foundation-model lineage as a free-text vs controlled-vocab field — leaning controlled (scGPT / Geneformer / scFoundation / scBERT / UCE / scPRINT / independent) so we can do "show me all Geneformer-line tools" filters later. Open.
- Atlas-platform tools (CELLxGENE Discover, HCA portal, CZ CELLxGENE Census) — these are tools but not in the standard "package on Bioc/PyPI" sense. Proposed: separate badge "Status: platform / atlas-infrastructure" that swaps the package row for an "Access" row pointing to a web portal.
- Lineage-tracing tools (DNA Typewriter, recording lineage, Gagnon-line) — these are wet-lab methods primarily, with analytical tooling secondary. Proposed: "Status: experimental method + companion analysis package" — two rows for source links (wet-lab protocol + analysis code).