Skip to main content
Back to top
Ctrl
+
K
UniVI
Getting started
Installation
Quickstart: integrate paired RNA + ATAC
Learn
Tutorials
Quickstart: integrate paired RNA + ATAC
CITE-seq: RNA + protein
Map query data onto a reference
Cell-type heads and supervised refinement
Imputation, denoising, generation, and perturbation
Custom modalities and likelihoods
Extended and experimental tutorials
In-silico chromatin perturbation of transcription regulators (ATAC → RNA)
Transformer encoders and a fused multimodal transformer
A unified RNA + protein + chromatin PBMC atlas from three assays
How much should you trust a cross-modal prediction?
Quality control from cross-modal disagreement: mis-paired cells and doublets
User guide
Preparing data
How the model works
Training and tuning
Using a trained model
Evaluating an integration
Advanced features
FAQ
Genome Research
Genome Research paper reproduction
Datasets
Paper analyses with the public API
Figs. 2–3 and S1: paired CITE-seq
Fig. 4 and S2: paired 10x Multiome (RNA + ATAC)
Fig. 5 and S5: bridging unpaired RNA and ATAC cohorts
Fig. 6 and S6: trimodal TEA-seq with a held-out well
Fig. 7 and S7: AML mosaic integration with mutation heads
Supplemental Fig. S3: SHARE-seq mouse skin
Supplemental Fig. S4: scNMT-seq mouse gastrulation (RNA, CpG, GpC)
Reference
API reference
univi.UniVIConfig
univi.ModalityConfig
univi.TrainingConfig
univi.ClassHeadConfig
univi.config.TokenizerConfig
univi.config.TransformerConfig
univi.refinement.RefinementConfig
univi.UniVIMultiModalVAE
univi.trainer.UniVITrainer
univi.refinement.UniVIRefiner
univi.preprocessing.RNAPreprocessor
univi.preprocessing.ADTPreprocessor
univi.preprocessing.ATACPreprocessor
univi.preprocessing.split_by_label
univi.workflows.make_loader
univi.data.MultiModalDataset
univi.data.align_paired_obs_names
univi.data.collate_multimodal_xy_recon
univi.datasets.list_datasets
univi.datasets.load
univi.datasets.fetch
univi.datasets.dataset_info
univi.datasets.register_dataset
univi.datasets.export_dataset
univi.datasets.get_data_dir
univi.datasets.pbmc_multiome_10k
univi.datasets.hao_citeseq_pbmc
univi.datasets.scnmt_gastrulation
univi.datasets.load_scnmt_gastrulation_genebody_triplet
univi.datasets.build_univi_inputs_from_scnmt_triplet
univi.evaluation.encode_adata
univi.evaluation.encode_fused_adata_pair
univi.workflows.stack_embeddings
univi.evaluation.cross_modal_predict
univi.evaluation.denoise_adata
univi.evaluation.generate_from_latent
univi.evaluation.fit_label_latent_gaussians
univi.evaluation.sample_latent_by_label
univi.evaluation.encode_moe_gates_from_tensors
univi.refinement.predict_heads_adata
univi.perturbation.predict_feature_perturbation
univi.evaluation.evaluate_alignment
univi.evaluation.compute_foscttm
univi.evaluation.compute_match_recall_at_k
univi.evaluation.compute_modality_mixing
univi.evaluation.compute_modality_entropy
univi.evaluation.label_transfer_knn
univi.evaluation.reconstruction_metrics
univi.evaluation.evaluate_cross_reconstruction
univi.evaluation.pearson_corr_per_feature
univi.evaluation.mse_per_feature
univi.evaluation.add_lsc17_scores
univi.workflows.save_reference
univi.workflows.load_reference
univi.utils.io.save_checkpoint
univi.utils.io.load_checkpoint
univi.utils.io.restore_checkpoint
univi.utils.seed.set_seed
univi.plotting.set_style
univi.plotting.umap
univi.plotting.umap_by_modality
univi.plotting.plot_confusion_matrix
univi.plotting.write_gates_to_obs
univi.plotting.plot_moe_gate_summary
univi.plotting.compare_raw_vs_pred_umap_features
univi.plotting.compare_raw_vs_denoised_umap_features
univi.plotting.plot_reconstruction_error_summary
univi.plotting.plot_featurewise_reconstruction_scatter
Changelog
Citation
Contributing
Repository
Suggest edit
Open issue
.rst
.pdf
univi.utils.seed.set_seed
Contents
set_seed()
univi.utils.seed.set_seed
#
univi.utils.seed.
set_seed
(
seed
=
0
,
deterministic
=
False
)
[source]
#
Contents
set_seed()