Mobile Menu

New biomarkers predicting COVID-19 severity discovered with multi-omics and machine learning

New multi-omics research has discovered biomarkers that can predict COVID-19 severity in patients.

The study, published in The Lancet, coupled multi-omics analysis with a machine-learning-based stratification. The results revealed an early signature of COVID-19 severity that was validated in a held-out dataset. 

Researchers from the Mayo Clinic in Rochester USA believe this is the largest profiling of plasma samples from a COVID-19 positive patient population using an integrative multi-omics approach.


COVID-19 is a multi-system disorder with high variability in clinical outcomes among patients who are admitted to hospital.

Although some cytokines such as interleukin (IL)-6 are believed to be associated with severity, there are no early biomarkers that can reliably predict patients who are more likely to have adverse outcomes.

“The ongoing COVID-19 pandemic is a public health emergency that would benefit from a systematic investigation of molecular alterations in patients to predict outcomes for improving clinical care,” noted the authors.

Patient samples

The team conducted a retrospective cohort study. They analysed samples from 455 patients with COVID-19 who had had a positive SARS-CoV-2 RT-PCR result and assigned participants to three subgroups depending on disease severity. A control cohort of 182 individuals was also assessed.

Next, they performed a deep profiling of circulatory cytokines and other proteins, lipids and metabolites across the samples. Most patient samples were collected prior to, or close to, hospital admission, representing ideal samples for predictive biomarker discovery.

The researchers then applied proximity extension assays to quantify cytokines and circulatory proteins and tandem mass spectrometry to measure lipids and metabolites.

Biomarker discovery

The team identified 1,463 cytokines and circulatory proteins, as well as 902 lipids and 1,018 metabolites.

They then developed a machine-learning-based prediction model. Machine learning approaches can be applied to highly complex and multidimensional profiling data to reveal novel signatures of clinical use. They discovered 102 biomarkers, which predicted severe and clinical COVID-19 outcomes better than the traditional set of cytokines. 

These predictive biomarkers included several novel cytokines and other proteins, lipids and metabolites. For example, altered amounts of C-type lectin domain family 6 member A (CLEC6A), ether phosphatidylethanolamine (P-18:1/18:1) and 2-hydroxydecanoate, had not previously been associated with severity in COVID-19.

Global proteomics and glycoproteomics were completed on a subset of patient samples with matched pre-COVID-19 plasma samples. This paired analysis examined how various molecules were altered in these individuals following infection. 

Reliably predicting COVID-19 severity

The absence of validation in an independent cohort remains a limitation of the study and future research is required in additional samples.

Nevertheless, this large study has countered the limitations seen in previous research. Preceding studies have suffered from small sample sizes (<100 participants), plasma samples collected at suboptimal times and group comparisons without adequate controls or without mild COVID-19 cases.

Availability of deep, unbiased profiling data using multi-omic platforms is crucial for the discovery of predictors of disease severity. Along with existing markers, these novel markers could lead to better management of patients and reduce mortality of COVID-19.

The authors reported, “A multiomic molecular signature in the plasma of patients with COVID-19 before being admitted to hospital can be exploited to predict a more severe course of disease.”

Written by Poppy Jayne Morgan, Front Line Genomics

Image Credit: Canva