Single-omics analyses are often the starting point for exploring specific molecular layers and achieving robust characterization of defined biological changes. However, biological systems are inherently complex and often require a broader analytical perspective to support a more comprehensive interpretation of underlying mechanisms, functional responses, and system-level changes. Multi-omics analyses provide this broader view by integrating complementary molecular layers and connecting signals across different biological levels.
As omics sciences continue to advance and research questions become increasingly complex, multi-omics integration has become an important step in obtaining a more comprehensive view of biological systems. In this context, the latest evolution of oloMAP Portal, our proprietary platform for the visualization and interpretation of omics data, has been the incorporation of functionalities for multi-omics integration.
Last Tuesday, November 18, our team of researchers and bioinformaticians, formed by Paolo Bonini (CEO), Lorenzo Torreni (CSO), Sajjan Singh Mehta (CTO) and Neus Pou Amengual (Bioinformatician), presented our new multi-omics analysis tool in a seminar open to the scientific community. The event took place at Barcelona Scientific Park and was also streamed online, bringing together more than 70 participants. The presentation covered both the technical aspects of multi-omics integration methods and their potential research applications and biological interpretation. Following the seminar, attendees had the opportunity to continue the discussion with the speakers and the rest of the oloBion team during a coffee break.
Reliable multi-omics studies typically benefit from clear conditions defined from the outset. First, the experimental design must be built around a clearly defined biological question, an adequate sample size to support the selected statistical approach, and a suitable selection of omics and assays. In parallel, high-quality omics data is critical for downstream interpretation, including well-annotated datasets generated through robust analytical methods and supported by appropriate preprocessing and statistical filtering steps.
Multi-omics integration can be approached through different methods, depending on the biological questions being addressed and the structure of the datasets. In our new multi-omics analysis tool, the data follow a vertical integration strategy, defined as the combination of multiple omics analyses performed on the same group of samples. The aim is to incorporate the distinct knowledge found in different levels of biochemical interactions in order to discover relationships across biological layers and support deeper biological interpretation. The analytical methods available on the platform are grouped into three integration levels:
- Low-Level/Early Integration: Datasets are combined without transformation or standardization to provide an overview of global patterns across omics layers. The methods included at this level are Correlation, which supports identification of associations between variables, and Pathway Analysis, which identifies biological pathways significantly enriched from a set of analytes, including metabolites, proteins and genes.
- Mid-Level/Middle Integration: Datasets are transformed and processed to improve comparability across omics layers. The statistical methods included at this level, Two- and N-omics, come from mixOmics, an R package that offers a wide range of multivariate methods for the exploration and integration of biological datasets, with a specific focus on variable selection.
- High-Level/Late Integration: Datasets are integrated into a unified representation through model-based approaches. The approach included at this level is machine learning, including classification and regression methods, which help to identify key molecular features and support predictive modelling.
The platform also includes a settings section that allows data filtering and preprocessing to be adapted to the analytical requirements of each omics type.
The extension of oloMAP Portal and the development of the multi-omics analysis tool reflect an important evolution in our analytical capabilities for multi-omics studies, opening new possibilities for connecting molecular layers and supporting deeper biological interpretation.
We would like to thank everyone who joined us, whether onsite or via livestream, to explore how integrated omics can help researchers build more coherent biological narratives. We would also like to express our gratitude to 𝘁𝗵𝗲 𝗡𝘂𝘁𝗿𝗶𝗮𝗹𝗶𝘁𝗲𝗰 𝗽𝗿𝗼𝗷𝗲𝗰𝘁 for supporting this initiative and for being part of such a meaningful and insightful discussion.