- July 12, 2021
- Posted by: Aanchal Iyer
- Category: Artificial Intelligence
Gene Expression Data to Detect Covid 19
The current outbreak of coronavirus disease (Covid-19) has affected millions of people while causing overwhelming mortality globally. Moreover, it should be noted that a cytokine storm during the viral infection has become an essential cause for the rising mortality rates.
Let us learn more about these cytokine storms and how gene expression data can help, read on.
AI Based Gene Expression Data
A team of researchers has used an Artificial Intelligence (AI) algorithm to examine terabytes of gene expression data to identify matching patterns in patients with past pandemic viral infections that include SARS, MERS, and swine flu.
The team, also includes Pradipta Ghosh from the University of California San Diego, indicates two tell-tale signatures. One, which is a set of 166 genes, discloses how the human immune system reacts to viral infections. The second set of 20 signature genes forecasts the severity of the disease – for example, the need to use a mechanical ventilator or to hospitalize. The algorithm’s utility was validated using lung tissues that were collected at autopsies from patients who died due to Covid-19 and animal models of the infection.
“These viral pandemic-associated signatures disclose how a person’s immune system reacts to a viral infection and how complicated and severe it might get and that gives us a road map for this as well as future pandemics,” states Ghosh.
Cytokine Storms During Viral Infections
Once a viral infection attacks a body, the immune system releases small proteins termed as cytokines into the blood. These proteins lead the immune cells to the site of infection to help fight the infection.
However, sometimes the body releases too many cytokines, which creates a runaway immune system which in turn attacks healthy tissues, indicates the study published in the journal biomedicine.
This mishap, termed as a cytokine storm, is supposed to be one of the main reasons why some patients infected by the virus, including some with the common flu, succumb to the infection while others do not. The data and information used to test and train the algorithm has been accessed from publicly available sources of patient gene expression data — all the RNA transcribed from patients’ genes and identified in blood samples or tissue.
Each time a new set of data from Covid-19 patients was made available, the researching team tested that very same data on their model and the same signature gene expression patterns have been observed each time. By examining the function and source of those genes in the first signature gene set, the study has also discovered the source of the cytokine storms — white blood cells known as T cells and macrophages and the cells lining lung airways.
Moreover, the results highlighted the results of the cytokine storms, that is, damage to those same lung airway cells and natural killer cells. This specialised immune cell that kills virus-infected cells.
The researchers believe that this information may also help guide treatment approaches for patients who are experiencing a cytokine storm by offering cellular targets and benchmarks to measure improvement. To test this theory, the team pre-treated rodents with either SARS-CoV-2-neutralising antibodies or a precursor version of Molnupiravir (a drug currently in clinical trials for the treatment of Covid-19).
After exposure to the Covid-19 virus, the lung cells of control-treated rodents showed the pandemic-associated 166- and 20-gene expression signatures, while the rodents who were pre-treated did not show those genes. This suggests that the treatments were effective in preventing a cytokine storm.