Yes. Taguchi, a professor at Chuo University, looks at a COVID-19 infection model that uses the blood of human patients who have been infected with COVID-19
From 2020 onwards, the COVID-19 pandemic started taking shape in almost all countries. Although vaccines currently appear to be effective in reducing COVID-19 mortality, the virus continues to mutate and the likelihood of another lockdown starting to rise increases. To avoid these situations, we certainly need effective drugs that have not yet been developed and an effective COVID-19 infection model.
In our previous articles published in Open Access Government Publication [1,2] We began our recent efforts to develop effective drugs for COVID-19 using computers.
However, our studies described in previous articles can only use human and mice cell lines. If we can use direct measurements using human patients infected with COVID-19, we may be able to get better results.
Using human cell lines to understand COVID-19
Recently, research groups led by Assistant Professor Miyata, Ryukyu University and Professor Ikematsu, National Institute of Technology, Okinawa College, in collaboration with us, employed our methods to analyze the gene expression of blood obtained from human COVID-19 patients . ,
This study has both advantages and disadvantages compared to the studies described in previous manuscripts. [1,2], Since this is a direct measurement from human patients, the measurement is more direct than those that use cell lines.
However, since it is not taken from the lungs, where the infection occurs, but rather from the blood, it is indirect in this sense. Thus, it is unclear whether replacement of human lung cell lines with human blood can improve outcome. The only way to understand this point is through a practical test.
Practical testing of gene datasets
The research team downloaded two sets of publicly available datasets, and applied our method to what they named PCAUFE.
They found that 123 genes were found to be differentially expressed between healthy controls and COVID-19 patients in the first data set. Since the total number of human genes is 20,000, 123 genes are very limited and constitute a small part of them.
To confirm whether this seemingly small, multiple gene has the potential to differentiate COVID-19 patients from healthy controls, the research group used three machine learning methods to classify two groups, patients and healthy controls. constructed the model, using only the selected 123 genes; Three models were tested using the second public data set, independent of the first data set.
To validate the efficiency of classification performance, the research group employed AUC, which takes 1.00 for correct performance and 0.5 for random selection. Three models trained by 123 genes can achieve an AUC greater than 0.9, which means excellent performance. Although the same process is repeated with the exchange of two data sets, that is, the model is trained with the second data set and tested with the first data set, it can achieve similar performance. This means the results are strong. Thus, despite the very small number of genes selected, they can successfully discriminate COVID-19 patients from healthy controls.
In addition, to confirm the superiority of PCAUFE, the research group also employed other state-of-the-art methods to select genes that are differentially expressed between COVID-19 patients and healthy controls. . Although the classification performance using genes selected by state-of-the-art methods is comparable to that of PCAUFE, when only the top-ranked equal number of genes selected by PCAUFE are used. Whereas the number of tests selected by state-of-the-art methods ranges from several thousand to eighteen thousand. Thus, state-of-the-art methods have little ability to limit the number of genes used for classification.
Enriching 123 genes
Next, the research group examined what types of functions are enriched in the selected 123 genes. They then found that the expression of several immune-related genes involved in these 123 genes were downregulated in the blood of COVID-19 patients. Furthermore, several biological pathways and transcription factors enriched in these genes were previously reported to be suppressed in COVID-19 patients.
These suggest that not only can PCAUFE identify genes whose expression can discriminate between COVID-19 patients and healthy controls (ie, biomarkers), but it can also identify a limited number of potentially disease-causing genes. can do
The discovery that blood samples from patients can be used for COVID-19 disease screening, forming a COVID-19 infection model, is noteworthy.
First, if there is no lung tissue, but blood can be an effective tissue for examination, it is much easier to collect. Collecting large numbers of lung samples from COVID-19 is frustrating, but collecting blood samples is possible. Since blood samples can be used for diagnosis, it is easier to monitor disease progression, which enables us to find time to treat with drugs if identified.
Unfortunately, the research team has not yet begun to identify potential drug candidate compounds using the 123 genes identified, this will be done soon, and they may yield promising candidate drug compounds.
 Yes. Taguchi, How to combat COVID-19 with a computer? Open Access Government, Issue 33, Jan (2022) pp. 210-211.
 Yes. Taguchi, can rats be an effective model animal for COVID-19? Open Access Government, Issue 34, Apr (2022) pp.112-113.
 Fujisawa, K., Shimo, M., Taguchi, YH. and others. PCA-based unsupervised feature extraction for gene expression analysis of COVID-19 patients. Science Rep. 11, 17351 (2021).
PLEASE NOTE: THIS IS A BUSINESS PROFILE
© 2019. This work is licensed under CC-BY-NC-ND.