Skip to main content

SpiceMix

overview

SpiceMix is an unsupervised tool for analyzing data of the spatial transcriptome. SpiceMix models the observed expression of genes within a cell as a mixture of latent factors. These factors are assumed to have some spatial affinity between neighboring cells. The factors and affinities are not known a priori, but are learned by SpiceMix directly from the data, by an alternating optimization method that seeks to maximize their posterior probability given the observed gene expression. In this way, SpiceMix learns a more expressive representation of the identity of cells from their spatial transcriptome data than other available methods.

SpiceMix can be applied to any type of spatial transcriptomics data, including MERFISH, seqFISH, HDST, and Slide-seq.

Install

pip install spicemix

Usage

from pathlib import Path

import anndata as ad
import torch

from spicemix.model import SpiceMixPlus

# Load datasets
datasets = []
replicate_names = []
for fov in range(5):
    dataset = ad.read_h5ad(f"./example_st_dataset_fov_{replicate}.h5ad") # Each dataset must have spatial information stored as an adjacency matrix
    name = f"{fov}"
    datasets.append(dataset)
    replicate_names.append(name)

# Define hyperparameters
K = 20 # Number of metagenes
lambda_Sigma_x_inv = 1e-4 # Spatial affinity regularization hyperparameter
torch_context = dict(device='cuda:0', dtype=torch.float32) # Context for PyTorch tensor instantiation 

# Initialize
spicemixplus_demo = SpiceMixPlus(
    K=K,
    datasets=datasets,
    lambda_Sigma_x_inv=lambda_Sigma_x_inv,
    torch_context=torch_context
)
    
# Train

## Initialization with NMF
for iteration in range(10):
    spicemixplus_demo.estimate_parameters(update_spatial_affinities=False)
    spicemixplus_demo.estimate_weights(use_neighbors=False)

## Using spatial information
num_iterations = 200
for iteration in range(num_iterations):
    spicemixplus_demo.estimate_parameters()
    spicemixplus_demo.estimate_weights()

# Save to disk
result_filepath = Path(f"./demo_{num_iterations}_iterations.h5ad")
spicemixplus_demo.save_results(result_filepath)
    
# Plot results

...

Tests

To run the provided tests and ensure that SpiceMix can run on your platform, follow the instructions below:

  • Download this repo.
git clone https://github.com/alam-shahul/SpiceMixPlus.git
  • Install pytest in your environment.
pip install pytest
  • Navigate to the root directory of this repo.
  • Run the following command. With GPU resources, this test should execute without errors in ~2.5 minutes:
python -m pytest -s tests/test_spicemix_shared.py

Cite

Cite our paper:

@article{chidester2020spicemix,
  title={SPICEMIX: Integrative single-cell spatial modeling for inferring cell identity},
  author={Chidester, Benjamin and Zhou, Tianming and Ma, Jian},
  journal={bioRxiv},
  year={2020},
  publisher={Cold Spring Harbor Laboratory}
}

paper

Metadata

Release files for spicemix 0.0.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for spicemix 0.0.1
File Size Uploaded Provenance Actions
spicemix-0.0.1.tar.gz 38.9 kB Without provenance Details for spicemix-0.0.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for spicemix 0.0.1
File Interpreter ABI Platform Actions
spicemix-0.0.1-py3-none-any.whl Python 3 none any Details for spicemix-0.0.1-py3-none-any.whl

Total release size: 85.5 kB

Release files / spicemix-0.0.1.tar.gz

Download URL spicemix-0.0.1.tar.gz
Size 38.9 kB
Tags Source
SHA-256 checksum
How to use checksums
5cd1f41f687b29b26ca20a597089dea34cd547a9c3a07669a477df693617438e
BLAKE2b-256 checksum
How to use checksums
02c0de17c91b344275dea2ae658fe709071c69f7cea5bab7efe10e018780d83f
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via python-httpx/0.23.0

Release files / spicemix-0.0.1-py3-none-any.whl

Download URL spicemix-0.0.1-py3-none-any.whl
Size 46.6 kB
Tags Python 3
SHA-256 checksum
How to use checksums
ad8b8cd9d138d132483f1d6bda54a1d376710e5732a551cc99e0a5299d0b39d7
BLAKE2b-256 checksum
How to use checksums
cf37505becc3c9af67e42915e8e2afe6d910b109d76d505ceac778d58f486c3b
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via python-httpx/0.23.0

Security for spicemix 0.0.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.0.1 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page