Transmorph: a unifying computational framework for modular single-cell RNA-seq data integration.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) data integration is essential for combining datasets from multiple experiments.
• Effective integration addresses batch effects and improves downstream analyses like clustering and cell type inference.
• Existing scRNA-seq integration tools often lack flexibility, hindering adaptation to specific research needs.

Purpose of the Study:

• To introduce the Transmorphic framework for flexible and powerful scRNA-seq data integration.
• To provide a software ecosystem that allows users to engineer custom data integration pipelines.
• To demonstrate the utility of Transmorphic in addressing practical scRNA-seq data challenges.

Main Methods:

• Development of the Transmorphic framework, an open-source Python package.
• Engineering of data integration pipelines using the Transmorphic framework.
• Application of Transmorphic to solve challenges in joint dataset embedding, gene space integration, and annotation transfer.

Main Results:

• Transmorphic enables the creation of adaptable scRNA-seq data integration pipelines.
• The framework successfully integrates datasets from different sources into a common representation.
• Demonstrated effectiveness in joint embedding, gene space integration, and cell cycle phase annotation transfer.

Conclusions:

• Transmorphic offers a flexible and powerful solution for scRNA-seq data integration challenges.
• The framework's software ecosystem supports the development of customized integration strategies.
• Transmorphic enhances scRNA-seq data analysis by facilitating robust integration across diverse datasets.