MOJITOO: a fast and universal method for integration of multimodal single-cell data

Abstract

Motivation The advent of multi-modal single-cell sequencing techniques have shed new light on molecular mechanisms by simultaneously inspecting transcriptomes, epigenomes and proteomes of the same cell. However, to date, the existing computational approaches for integration of multimodal single-cell data are either computationally expensive, require the delineation of parameters or can only be applied to particular modalities. Results Here we present a single-cell multi-modal integration method, named Multi-mOdal Joint IntegraTion of cOmpOnents (MOJITOO). MOJITOO uses canonical correlation analysis for a fast and parameter free detection of a shared representation of cells from multimodal single-cell data. Moreover, estimated canonical components can be used for interpretation, i.e. association of modality-specific molecular features with the latent space. We evaluate MOJITOO using bi- and tri-modal single-cell datasets and show that MOJITOO outperforms existing methods regarding computational requirements, preservation of original latent spaces and clustering. Availability and implementation The software, code and data for benchmarking are available at https://github.com/CostaLab/MOJITOO and https://doi.org/10.5281/zenodo.6348128. Supplementary information Supplementary data are available at Bioinformatics online.

Ocelli: an open-source tool for the visualization of developmental multimodal single-cell data

“…2). While some benchmarking strategies have been developed for assessing analysis and visualization of well-clustered data, such as silhouette score or adjusted Rand index score [48], we designed new benchmarks related to the visualization of developmental multimodal data and transitions between cells (Fig. 3, Supplementary Fig.…”

“…It uses stochastic variational inference [46, 47] for scalability of the parameter inference. Other approaches based on integrative methods exist, such as MOJITOO [48] that uses canonical correlation analysis to find low-dimensional multimodal data representation.…”

scMEGA: single-cell multi-omic enhancer-based gene regulatory network inference

“…SHARE-seq ( Ma et al , 2020 ) and 10× Multiome], this integration step can be skipped. Here, users only need to create a joint embedding of cells by using for example the CCA integration from MOJITOO ( Cheng et al , 2022 ).…”

scMEGA: Single-cell Multiomic Enhancer-based Gene Regulatory Network Inference

“…Finally, scMEGA merges the data for each of the pairs by extracting gene expression from scRNAseq and chromatin accessibility from scATAC-seq. Of note, this step can be skipped in case single-cell multimodal protocols (e.g., SHARE-seq [16] and 10X Multiome) were employed and a joint embedding of the modalities is available, such as provided by MOJI-TOO [17].…”