How Multi-omics Studies are Unveiling New Insights into Complex Diseases

In the dynamic landscape of scientific research, multi-omics approaches have emerged as a transformative force, revolutionizing the way researchers investigate complex biological systems. By concurrently analyzing genomics, proteomics, metabolomics, lipidomics, and other ‘omics’ layers, scientists have been able to unravel intricate molecular networks and gain a more holistic understanding of biological systems, especially in the context of disease. Each omics layer can provide important insight on its own, but through multi-dimensional analyses we can better understand the dynamic interplay of genes, proteins, and small molecules in biological processes.

With the rise of high-throughput bioanalytical technologies, enhanced automation, and improved AI and machine learning tools, researchers can more seamlessly and simultaneously generate, integrate, and analyze these datasets to identify meaningful correlations, develop new hypotheses, and drive groundbreaking discoveries that inform better interventions and development of effective therapeutics.

Here we touch on a few recent examples showcasing how multi-omics research has “moved the needle” in expanding our understanding of complex diseases.

Unraveling the Biology of COVID-19

This March marks four years since COVID-19 was declared a global pandemic. Scientists around the world were quick to respond, utilizing the latest innovative technologies to understand not only the genetics and structure of the SARS-CoV-2 virus, but also its pathogenesis and immune response signatures. These integrative approaches led to the creation of multiple vaccine candidates within the same year and continue to be important tools for quickly characterizing emerging variants, monitoring spread, developing new vaccines and therapeutics, and informing public health policies.

While genomics and transcriptomics enabled identification of SARS-CoV-2 mutations, investigating circulating proteins and metabolites can help elucidate systemic mechanisms underlying COVID-19. This review paper by Michele Costanzo et. al., titled “COVIDomics: The Proteomic and Metabolomic Signatures of COVID-19,” summarizes key findings from multiple proteomic and metabolomic studies of COVID-19 in both plasma and serum. For example, the paper references the Su et al. study which identified a major shift in patients’ omics signatures between mild and moderate disease, as seen by an increase in pro-inflammatory cytokines and immune cell activation proteins in combination with altered amino acids, nucleotides, and carbohydrates. Costanzo et. al. also performed their own bioinformatic analyses integrating data from the papers to further elucidate potential pathways of interest. From the proteomics analysis they found clusters of proteins related to immune responses, platelet degranulation, and lipid transport. From the metabolomics side, the group identified dysregulations related to several amino acid metabolic pathways as well as energy metabolism.

Beyond acute COVID-19 infection, multi-omics-based approaches are also being used to better understand and develop new treatment strategies for long COVID. Also known as post-acute sequelae of SARS-CoV-2 infection (PASC), patients with long COVID experience symptoms such as fatigue, shortness of breath, chronic pain, and trouble concentrating. These symptoms range from mild to debilitating and can severely impact quality of life.

Kaiming Wang et. al. recently utilized a multi-omics strategy to identify clinical phenotypes and predictive biomarkers associated with long COVID. The group obtained plasma samples from 117 individuals during acute phase infection and 6 months post-infection to profile and assess changes in cytokines as well as the broader proteome and metabolome. They found that 231 molecules were significantly altered between acute infection and healthy control samples (24 cytokines, 63 proteins, and 144 metabolites) and 157 differentially expressed molecules between the acute infection and post-infection (eight cytokines, 34 proteins, and 115 metabolites). Their network analysis revealed a sustained inflammatory response, platelet degranulation, and cellular activation post-infection as well as dysregulation in arginine biosynthesis, methionine metabolism, taurine metabolism, and tricarboxylic acid (TCA) cycle processes.

Going a step further, Wang et. al. developed a prognostic model encompassing 20 molecules (seven cytokines and 13 metabolites) involved in regulating T cell exhaustion and energy metabolism that can reliably predict adverse clinical outcomes following discharge from initial COVID infection with 83% accuracy and an AUC of 0.96.

Characterizing Potential Molecular Mechanisms Underlying Schizophrenia

Schizophrenia was first described in the early 1900s, yet the biology surrounding this disease has still been challenging to discern. Many schizophrenia symptoms overlap with other mental illnesses, making it difficult for clinicians to make a definitive diagnosis and provide the most appropriate treatment.

In a recent study titled, “Consolidation of metabolomic, proteomic, and GWAS data in connective model of schizophrenia”, Arthur Kopylov and colleagues integrated three omics layers with quantitative analysis to propose a map of molecular events associated with schizophrenia psychopathology.

Evaluating samples from 77 patients with schizophrenia and 61 healthy donors, the group was able to identify 20 differently expressed proteins, with almost half of them considered new for schizophrenia. Using metabolomic analyses they further identified 18 group-specific compounds, most of which were the part of transformation of tyrosine and steroids with the prevalence to androgens. Of note, the GWAS assay mostly failed to reveal significantly associated loci.

Among the significantly differed proteins, APOC3, PLMN, APOH, A1AG1, VTDB, APOB, ITIH4, and ITIH1 were the most meaningful proteins identified. The majority of these are associated with lipid metabolism (including cholesterol uptake and dependent steroidogenesis) and associated synthesis and transport of ligands acting in chronic inflammatory reactions. Combining the proteomics and metabolomics data, the authors were able to propose an integrative scheme depicting the interplay between proteome and metabolome layers in patients with schizophrenia, hypothesizing that immune response and lipid metabolism might be early affected pathways in this disease.

Identifying Pathways and Markers to Track Time-to-Delivery and Fetal Growth

Pre-term birth is the leading cause of death in children under five, particularly in low- and middle-income countries. Premature children also have an increased risk of both short- and long-term complications, including neurological, cardiovascular, and metabolic conditions.

In a recent study titled, “Multiomic signals associated with maternal epidemiological factors contributing to preterm birth in low- and middle-income countries,” authors Camilo Espinosa et. al. collected maternal medical history data and plasma samples during early to mid-pregnancy from a cohort of 13,841 women across four low- and middle-income countries. The group further analyzed plasma samples from a subcohort of 231 pregnant women to generate proteomic, metabolomic, and lipidomic datasets. From this, the team identified biological signatures within the multi-omics subcohort that could capture the time from sampling to delivery. In addition to pregnancy-associated steroid hormones, Espinosa et. al. observed increasing levels of immune system–associated proteins such as PD-L1, CCL28, and LIFR in maternal plasma closer to the onset of labor. From the metabolomics side, the linoleate metabolism pathway was also significantly correlated.

A recent University of Oxford study led by Jose Villar et. al. and supported by the Sapient team also identified putative metabolic pathways that influence fetal growth trajectories and childhood health outcomes. Sapient performed simultaneous metabolomics and lipidomics analyses on over 3,500 biological samples obtained from mothers in early pregnancy who were enrolled in the INTERBIO-21st Study. The data generated identified specific patterns of lipid metabolites that closely tracked with fetal growth trajectories and point to potential mechanisms that regulate fetal growth during development.

The integration of multi-omics approaches has undeniably transformed the landscape of biological research, offering unprecedented insights into the intricate molecular mechanisms underlying complex diseases. The examples discussed, ranging from the elucidation of COVID-19 pathogenesis to the identification of predictive biomarkers for long COVID, schizophrenia, and preterm birth, highlight the versatility and power of multi-omics analyses.

As technological advancements and data integration methods continue to evolve, researchers are poised to unlock even greater depths of understanding, facilitating the development of targeted therapies, personalized medicine, and improved public health strategies. The collaborative efforts across genomics, transcriptomics, proteomics, and metabolomics, supported by enhanced automation and AI/ML tools, signify a promising future where multi-omics approaches play a pivotal role in addressing complex challenges in drug development and in advancing our knowledge to improve human health.

In addition to metabolomics and lipidomics, Sapient now also offers high throughput discovery proteomics, further expanding the multi-omics insights that can be captured in every sample. You can schedule a time to talk to our team to learn more our proteomics capabilities and how these data can be integrated in comprehensive biocomputational analyses.