PULSAR

⚠️ Status: partial | Version: v1.0

Overview

Multi-scale multicellular biology modeling, captures cell-cell communication and tissue-level organization

!!! tip “When to choose PULSAR”

User wants cell-cell communication analysis, tissue-level modeling, multicellular programs, or intercellular signaling

Specifications

Property

Value

Model

PULSAR

Version

v1.0

Tasks

embed, integrate

Modalities

RNA

Species

human

Gene IDs

symbol

Embedding Dim

512

GPU Required

Yes

Min VRAM

16 GB

Recommended VRAM

32 GB

CPU Fallback

No

Adapter Status

⚠️ partial

Quick Start

import omicverse as ov

# 1. Check model spec
info = ov.fm.describe_model("pulsar")

# 2. Profile your data
profile = ov.fm.profile_data("your_data.h5ad")

# 3. Validate compatibility
check = ov.fm.preprocess_validate("your_data.h5ad", "pulsar", "embed")

# 4. Run inference
result = ov.fm.run(
    task="embed",
    model_name="pulsar",
    adata_path="your_data.h5ad",
    output_path="output_pulsar.h5ad",
    device="auto",
)

# 5. Interpret results
metrics = ov.fm.interpret_results("output_pulsar.h5ad", task="embed")

Input Requirements

Requirement

Detail

Gene ID scheme

symbol

Preprocessing

Standard preprocessing. For tissue-level analysis, include spatial or neighborhood context if available.

Data format

AnnData (.h5ad)

Batch key

.obs column for batch integration (optional)

Output Keys

After running ov.fm.run(), results are stored in the AnnData object:

Key

Location

Description

X_pulsar

adata.obsm

Cell embeddings (512-dim)

 
import scanpy as sc

adata = sc.read_h5ad("output_pulsar.h5ad")
embeddings = adata.obsm["X_pulsar"]  # shape: (n_cells, 512)

# Downstream analysis
sc.pp.neighbors(adata, use_rep="X_pulsar")
sc.tl.umap(adata)
sc.tl.leiden(adata, resolution=0.5)
sc.pl.umap(adata, color=["leiden"])

Resources

Hands-On Tutorial

For a step-by-step walkthrough with code, see the PULSAR Tutorial Notebook.