scgpt

# scGPT — Single-Cell Foundation Model

## Prerequisites

| Requirement | Minimum | Recommended |
| ----------- | ------- | ----------- |
| Python      | 3.10+   | 3.11        |
| CUDA        | 12.1+   | 12.4+       |
| GPU VRAM    | 16 GB   | 24 GB+      |

## How to run

### Loading the vocabulary and checkpoint

scGPT checkpoints are **raw directories** (`args.json`, `best_model.pt`,
`vocab.json`) — not Hugging Face hub repos. Point at the directory, not an HF
repo id.

```python
from scgpt.tokenizer.gene_tokenizer import GeneVocab
gv = GeneVocab.from_file("/path/to/scgpt-human/vocab.json")
print(len(gv))   # 60697 for the released human checkpoint
```

### Embedding an AnnData

```python
import anndata as ad
from scgpt.tasks import embed_data

adata = ad.read_h5ad("dataset.h5ad")        # var must contain a gene-name column
emb = embed_data(
    adata,
    model_dir="/path/to/scgpt-human",
    gene_col="feature_name",
    use_fast_transformer=False,             # see Gotchas
)
# emb is an AnnData with .obsm["X_scGPT"]
```

## Output format

`embed_data` returns an `AnnData` whose `.obsm["X_scGPT"]` is the per-cell
embedding (`n_cells × emb_dim`, 512 by default). Downstream: feed to
`scanpy.pp.neighbors` / `scanpy.tl.umap`.


## Remote compute

Needs ≥24 GB VRAM and the released human checkpoint (~200 MB:
`args.json`, `best_model.pt`, `vocab.json`). Read
`compute_details({provider, mode:'read'})` for an environment with `scgpt`
and a pre-cached checkpoint directory, then:

```python
c = host.compute.create(provider)
job = c.submit_job(
    intent="scGPT embed 50k cells — 1×GPU, ~5 min",
    inputs=[
        {"src": "dataset.h5ad", "dst_filename": "dataset.h5ad"},
        {"src": "embed.py", "dst_filename": "embed.py"},
    ],
    command="python3 embed.py",
    environment=...,   # env name from compute_details
    outputs=["embedded.h5ad"],
    timeout_seconds=1800,
)
print(job.job_id)   # cell ends here — kernel never blocks on compute
```

Then call the `wait_for_notification` brain-tool. When the
`compute_done` notification arrives, act on its payload:

```python
save_artifacts(payload["featured_files"])   # paths under hpc/<job_id>/
```

For the full result dict (`output_files`, `remote_workdir`, …), re-enter the
kernel and bind the compute handle separately — `.close()` lives on the
handle, not on the job object:

```python
h = host.compute.create(provider)
res = h.attach_job(job_id).result()
h.close()
```

See the `remote-compute-ssh` / `remote-compute-modal` skill for the
orchestration details.

In `embed.py`, pass `model_dir=` the checkpoint path from `compute_details`.
If `flash-attn` is unavailable in that environment, set
`use_fast_transformer=False`.


## Gotchas

- **`use_fast_transformer` default is `True`** but resolves to a FlashAttention
  path that may not import in every env. Pass `use_fast_transformer=False`
  unless you've confirmed `flash_attn` loads cleanly.
- The package historically depended on `torchtext.vocab.Vocab`; in
  environments without torchtext a pure-Python shim provides `Vocab` —
  functionally identical for `GeneVocab`, but if you hit
  `AttributeError: 'Vocab' object has no attribute …`, you're on a stale shim.
- Gene names must match the vocab; unmatched genes are dropped. Set
  `gene_col` to the column in `adata.var` that holds symbols.

## Troubleshooting

| Symptom                                           | Fix                                              |
| ------------------------------------------------- | ------------------------------------------------ |
| `flash_attn is not installed` warning at import   | Harmless; pass `use_fast_transformer=False`      |
| `'Vocab' object has no attribute 'vocab'`         | Env has an old torchtext shim — update the env   |
| Nearly all genes dropped                          | Wrong `gene_col`; check `adata.var.columns`      |
| "scgpt not in manifest" / env-detection misses scGPT | The baked env manifest lists the distribution as `scGPT` (and `flash_attn`), pip's canonical casing — normalize manifest keys before lookup: `name.lower().replace('-', '_')` |

---

**Next**: cluster/annotate the embedding with the scanpy library
(`sc.pp.neighbors` → `sc.tl.leiden` / `sc.tl.umap`), or compare to an
scvi-tools latent space on the same data.