scvi-tools — scVI / scANVI

scvi-tools (Gayoso et al. 2022, github.com/scverse/scvi-tools, BSD-3-Clause) wraps a family of deep generative models for single-cell omics. The scRNA-seq core is scVI (unsupervised batch-corrected latent embedding) and scANVI (scVI + a classifier head for semi-supervised cell-type label transfer). Both expect raw integer UMI counts and emit a low-dimensional X_scVI / X_scANVI that drops into the scanpy neighbors → leiden → umap pipeline.

How to run

scVI — batch-corrected latent space

import scanpy as sc
import scvi

adata = sc.read_h5ad("dataset.h5ad")
adata.layers["counts"] = adata.X.copy # preserve raw BEFORE any normalize/log1p
sc.pp.normalize_total(adata); sc.pp.log1p(adata) # optional, for HVG / plotting only
sc.pp.highly_variable_genes(adata, n_top_genes=2000, batch_key="batch", subset=True)

scvi.model.SCVI.setup_anndata(adata, layer="counts", batch_key="batch")
model = scvi.model.SCVI(adata, n_latent=30)
model.train(max_epochs=200, early_stopping=True, accelerator="gpu", devices=1)

adata.obsm["X_scVI"] = model.get_latent_representation
adata.layers["scvi_normalized"] = model.get_normalized_expression(library_size=1e4)

scANVI — label transfer from a partially-annotated reference

lvae = scvi.model.SCANVI.from_scvi_model(
    model, labels_key="cell_type", unlabeled_category="Unknown")
lvae.train(max_epochs=20, n_samples_per_label=100, accelerator="gpu", devices=1)

adata.obsm["X_scANVI"] = lvae.get_latent_representation
adata.obs["pred_cell_type"] = lvae.predict

accelerator="gpu", devices=1 is the PyTorch-Lightning spelling; the legacy use_gpu= kwarg was removed in scvi-tools 1.x and now raises TypeError.

Differential expression

de = model.differential_expression(
    groupby="leiden", group1="3", # group2=None → vs. all other cells
    mode="change", delta=0.25)
top = de.sort_values("proba_de", ascending=False).head(50)

For one-vs-rest leave group2 out — "rest" is scanpy's rank_genes_groups convention, not scvi-tools'; here group2 is a literal category name and "rest" would match zero cells.

scvi-tools ≥1.4 defaults to mode="vanilla", whose result columns are exactly:

['proba_m1', 'proba_m2', 'bayes_factor', 'scale1', 'scale2', 'raw_mean1',
 'raw_mean2', 'non_zeros_proportion1', 'non_zeros_proportion2',
 'raw_normalized_mean1', 'raw_normalized_mean2', 'comparison', 'group1',
 'group2']

— no lfc_*, no proba_de, no is_de_fdr_*. Pass mode="change" to get lfc_mean / lfc_median / proba_de / is_de_fdr_0.05. Sort on proba_de (or on bayes_factor if you deliberately stayed in vanilla mode).

Output format

Key	What
adata.obsm["X_scVI"]	n_cells × n_latent batch-corrected embedding
adata.obsm["X_scANVI"]	label-aware embedding (better separates known classes)
adata.obs["pred_cell_type"]	scANVI predicted label per cell
adata.layers["scvi_normalized"]	decoded expression, library-size normalized
DE dataframe	per-gene lfc_* / proba_de (with mode="change")

Remote compute

A100-class GPU recommended for >50k cells. The prebuilt singlecell_gpu NVIDIA NIM env ships scvi-tools 1.4.2 + scanpy 1.11.5 + anndata 0.11.4 — read compute_details({provider: 'byoc:nvidia', mode: 'read'}) for the current image ref, then:

c = host.compute.create('byoc:nvidia', provider_params={'nvidia': {
    'image': '<image ref from compute_details>', # e.g. im-...
    'gpu': 'A100',
    'cpu': 8,
    'memory': 32768,
    'timeout': 3600,
}})
job = c.submit_job(
    intent="scVI+scANVI on 80k cells — 1×A100, ~15 min",
    inputs=[
        {"src": "dataset.h5ad", "dst_filename": "dataset.h5ad"},
        {"src": "pipeline.py", "dst_filename": "pipeline.py"},
    ],
    command="python pipeline.py",
    outputs=["out/**"],
    timeout_seconds=2400)
print(job.job_id) # cell ends here — kernel never blocks on compute

h5ad_safe_obs is auto-loaded into the local analysis kernel only — in pipeline.py running on the NVIDIA NIM job, paste the helper at the top of the script (or inline the pd.Index(np.asarray(..., dtype=object)) coercion) before .write_h5ad.

Then poll from a later cell — .result() is one non-blocking probe, and it is what harvests the outputs once the job is terminal; nothing runs in the background, so a job you never poll is never harvested. Bind the compute handle (not the job) to poll from a fresh kernel — .close() lives on the handle, not on the job:

h = host.compute.create('byoc:nvidia')
res = h.attach_job(job_id).result() # status, output_files, remote_workdir,...
if res["status"] == "succeeded":    # else end the cell and poll again later
    for path in res["featured_files"]:
        host.save_artifact(path)
    h.close()

See the remote-compute-nvidia skill for orchestration details.

Gotchas

Gotcha	What happens / fix
differential_expression defaults to mode="vanilla" (scvi-tools ≥1.4)	KeyError: 'lfc_mean' / 'proba_de' when sorting — pass mode="change" to get lfc_*/proba_de/is_de_fdr_*; in vanilla mode sort on bayes_factor.
adata.obs index/columns are string[pyarrow] (ArrowStringArray)	.write_h5ad dies with IORegistryError: No method registered for writing <class 'pandas.arrays.ArrowStringArray'> (anndata #2377). Coerce before writing: adata.obs = h5ad_safe_obs(adata.obs) (kernel helper — local kernel only; inline the coercion in remote pipeline.py). .astype(str) alone is not enough — on a pyarrow-backed Index/Series it returns another Arrow-backed array; round-trip through np.asarray(..., dtype=object). anndata.settings.allow_write_nullable_strings = True does not cover Arrow-backed strings.
use_gpu= kwarg	Removed in 1.x → TypeError: train got an unexpected keyword argument 'use_gpu'. Use accelerator="gpu", devices=1.
Log-normalized data fed to setup_anndata	Silent garbage — scVI's NB likelihood needs raw integer counts. Stash counts in adata.layers["counts"] before normalize/log1p and pass layer="counts".

Troubleshooting

Symptom	Fix
KeyError: 'lfc_mean' (or 'proba_de', 'is_de_fdr_0.05') on DE result	Add mode="change" to differential_expression; the default vanilla mode has no LFC columns.
IORegistryError: No method registered for writing <class 'pandas.arrays.ArrowStringArray'> on .write_h5ad	adata.obs = h5ad_safe_obs(adata.obs) (and adata.var if needed) before writing. The allow_write_nullable_strings flag does not help here.
TypeError:... unexpected keyword argument 'use_gpu'	Replace with accelerator="gpu", devices=1.
ValueError:... non-negative integers / NB loss explodes	layer="counts" points at log/float data — restore raw counts.
MisconfigurationException: No supported gpu backend found	No CUDA visible — drop accelerator/devices to fall back to CPU, or dispatch via Remote compute.
UnicodeEncodeError: 'ascii' codec can't encode character... writing a summary / printing	Container has no LANG so Python defaults to ASCII. Open files with encoding="utf-8" and/or sys.stdout.reconfigure(encoding="utf-8") at script top. The prebuilt singlecell_gpu env sets PYTHONIOENCODING=utf-8, so this only bites user-built images.
AttributeError:... object has no attribute 'close' on a job handle	You chained host.compute.create(...).attach_job(...) and called .close on the job. Bind the compute handle separately and close that — see Remote compute above.

---

Next: cluster on X_scVI with scanpy (sc.pp.neighbors(use_rep="X_scVI") → sc.tl.leiden → sc.tl.umap); for spatial deconvolution train cell2location / DestVI / Tangram on the scRNA-seq reference.