cnmfsns

cNMF-SNS: powerful factorization-based multi-omics integration toolkit

Authors: Ted Verhey, Heewon Seo, Sorana Morrissy

cNMF-SNS (consensus Non-negative Matrix Factorization Solution Network Space) is a Python package enabling mosaic integration of bulk, single-cell, and spatial expression data between and within datasets. Datasets can have partially overlapping features (eg. genes) as well as non-overlapping features. cNMF provides a robust, unsupervised deconvolution of each dataset into gene expression programs (GEPs). Network-based integration of GEPs enables flexible integration of many datasets across assays (eg. Protein, RNA-Seq, scRNA-Seq, spatial expression) and patient cohorts.

Communities with GEPs from multiple datasets can be annotated with dataset-specific annotations to facilitate interpretation.

⚡Main Features

Here are just a few of the things that cNMF-SNS does well:

• Identifies interpretable, non-negative programs at multiple resolutions
• Mosaic integration does not require subsetting features/genes to a shared or overdispersed subset
• Ideal for incremental integration (adding datasets one at a time) since deconvolution is performed independently on each dataset
• Integration performs well even when the datasets have mismatched features (eg. Microarray, RNA-Seq, Proteomics) or sparsity (eg single-cell vs bulk RNA-Seq and ATAC-Seq)
• Two interfaces: command-line interface for rapid data exploration and python interface for extensibility and flexibility

🔧 Install

☁️ Public Release

Install the package with conda (in an isolated conda environment)

conda create -n cnmfsns -c conda-forge cnmfsns
conda activate cnmfsns

📖 Documentation

🗐 Data guidelines

cNMF-SNS can factorize a wide variety of datasets, but will work optimally in these conditions:

• Use untransformed (raw) data where possible, and avoid log-transformed data.
• For single-cell or spatial RNA-Seq data, the best data to use is feature counts, then TPM-normalized values, then RPKM/FPKM-normalized values.

📓 Python interface

To get started, sample proteomics datasets and a Jupyter notebook tutorial is available here.

Detailed API reference can be found on ReadTheDocs.

⌨️ Command line interface

See the command line interface documentation.

💭 Getting Help

For errors arising during use of cNMF-SNS, create and browse issues in the GitHub "issues" tab.