Abstract
Multi-omics approaches use a diversity of high-throughput technologies to profile the different molecular layers of living cells. Ideally, the integration of this information should result in comprehensive systems models of cellular physiology and regulation. However, most multi-omics projects still include a limited number of molecular assays and there have been very few multi-omic studies that evaluate dynamic processes such as cellular growth, development and adaptation. Hence, we lack formal analysis methods and comprehensive multi-omics datasets that can be leveraged to develop true multi-layered models for dynamic cellular systems. Here we present the STATegra multi-omics dataset that combines measurements from up to 10 different omics technologies applied to the same biological system, namely the well-studied mouse pre-B-cell differentiation. STATegra includes high-throughput measurements of chromatin structure, gene expression, proteomics and metabolomics, and it is complemented with single-cell data. To our knowledge, the STATegra collection is the most diverse multi-omics dataset describing a dynamic biological system.
Introduction: The Need for Integrative Systems Biology
In the era of "big data" biology, researchers have become adept at profiling individual molecular layers—such as the transcriptome (all expressed genes) or the proteome (all present proteins). However, biological processes are inherently multi-layered and dynamic; interactions between these layers drive cellular physiology. A significant challenge in the field is that most existing multi-omics datasets are either static snapshots or restricted to only two or three molecular layers. This limitation hinders the development of true systems-level models capable of explaining how regulatory signals propagate from chromatin accessibility through RNA and protein levels to eventually alter cellular metabolism.
The STATegra project addresses this gap by providing a deeply integrated, time-resolved dataset of a well-characterized biological process: the differentiation of mouse pre-B cells. This "workbench" serves as a gold-standard resource for the community to develop and test integrative systems biology algorithms.
Experimental Design and Methods
The study utilized the B3 cell line, a robust model for mouse pre-B cell differentiation. To capture the dynamics of this process, the researchers employed an inducible system where cells were engineered with a version of the Ikaros transcription factor (Ikzf1-Ert2). This factor only translocates to the nucleus upon treatment with 4-hydroxytamoxifen (4-OHT), allowing researchers to precisely trigger the differentiation program.
By activating Ikaros, the team could "start the clock" on differentiation, harvesting samples at six precise time points: 0, 2, 6, 12, 18, and 24 hours. This temporal resolution is critical for observing the sequential regulatory events that define cellular transition.
The STATegra Multi-Omics Resource
What distinguishes STATegra is the unprecedented breadth of technologies applied to the same biological samples. The collection comprises 793 different samples spanning ten distinct molecular layers:
- Gene Expression: mRNA-seq and miRNA-seq.
- Chromatin Structure: DNase-seq (accessibility), RRBS (DNA methylation), and ChIP-seq (protein-DNA binding).
- Proteomics: Shotgun proteomics (LC-MS/MS).
- Metabolomics: Targeted LC-MS and GC-MS.
- Single-Cell Data: scRNA-seq and scATAC-seq.
All raw data has been deposited in public repositories (GEO, ProteomeXchange, MetaboLights), with pre-processed matrices available for immediate use by bioinformaticians via Lifebit and GitHub.
Key Insights and Integrative Results
The high-resolution nature of the dataset allowed for several key biological and methodological observations:
- Temporal Cascades: The data reveals that molecular layers respond at different speeds. Changes in chromatin accessibility and RNA levels generally precede alterations in proteins and metabolites, suggesting a predictable regulatory "lag" as signals flow through the cell.
- Regulatory Consistency: The integrity of the dataset was validated by tracking specific genes across all layers. For example, Ikaros binding at the Ldha promoter (ChIP-seq) correlated with decreased chromatin accessibility (DNase-seq), which subsequently led to lower mRNA and protein levels, and finally, a measurable decrease in lactate production. This "vertical" validation confirms that the dataset accurately captures the biological flow of information.
- High Replicability: Biological replicates showed extremely high correlation coefficients (>0.9), ensuring that the observed changes are biologically significant rather than experimental noise.
- Metabolic Switch: The multi-omics integration successfully captured the known metabolic transition in B-cells, where they shift away from anaerobic glycolysis as they differentiate.
Discussion and Future Directions
While the STATegra dataset is comprehensive, the authors acknowledge common hurdles in high-dimensional biology. Proteomics data, for instance, contained missing values inherent to LC-MS/MS technology, requiring sophisticated imputation. Additionally, batch effects were noted in the metabolomics arm, and single-cell ATAC-seq was limited to the 0h and 24h time points.
Despite these challenges, STATegra stands as a unique "workbench" for the development of multi-omics integration methods. Future work will leverage this resource to train machine learning models capable of inferring complex regulatory networks across different cell types and species, moving us closer to a holistic understanding of cellular dynamics.
Summary updated