Multi-objective Evolutionary Discretization of Gene Expression Profiles: Application to COVID-19 Severity Prediction

Machine learning models can use information from gene expressions in patients to efficiently predict the severity of symptoms for several diseases. Medical experts, however, still need to understand the reasoning behind the predictions before trusting them. In their day-to-day practice, physicians prefer using gene expression profiles, consisting of a discretized subset of all data from gene expressions: in these profiles, genes are typically reported as either over-expressed or under-expressed, using discretization thresholds computed on data from a healthy control group. A discretized profile allows medical experts to quickly categorize patients at a glance. Building on previous works related to the automatic discretization of patient profiles, we present a novel approach that frames the problem as a multi-objective optimization task: on the one hand, after discretization, the medical expert would prefer to have as few different profiles as possible, to be able to classify patients in an intuitive way;on the other hand, the loss of information has to be minimized. Loss of information can be estimated using the performance of a classifier trained on the discretized gene expression levels. We apply one common state-of-the-art evolutionary multi-objective algorithm, NSGA-II, to the discretization of a dataset of COVID-19 patients that developed either mild or severe symptoms. The results show not only that the solutions found by the approach dominate traditional discretization based on statistical analysis and are more generally valid than those obtained through single-objective optimization, but that the candidate Pareto-optimal solutions preserve the sense-making that practitioners find necessary to trust the results. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Recent computational drug repositioning strategies against SARS-CoV-2.

Since COVID-19 emerged in 2019, significant levels of suffering and disruption have been caused on a global scale. Although vaccines have become widely used, the virus has shown its potential for evading immunities or acquiring other novel characteristics. Whether current drug treatments are still effective for people infected with Omicron remains unclear. Due to the long development cycles and high expense requirements of de novo drug development, many researchers have turned to consider drug repositioning in the search to find effective treatments for COVID-19. Here, we review such drug repositioning and combination efforts towards providing better handling. For potential drugs under consideration, aspects of both structure and function require attention, with specific categories of sequence, expression, structure, and interaction, the key parameters for investigation. For different data types, we show the corresponding differing drug repositioning methods that have been exploited. As incorporating drug combinations can increase therapeutic efficacy and reduce toxicity, we also review computational strategies to reveal drug combination potential. Taken together, we found that graph theory and neural network were the most used strategy with high potential towards drug repositioning for COVID-19. Integrating different levels of data may further improve the success rate of drug repositioning.

Integrated bioinformatics analysis identifies shared immune changes between ischemic stroke and COVID 19.

Although COVID-19 is primarily a respiratory disease, its neurological complications, such as ischemic stroke (IS), have aroused growing concerns and reports. However, the molecular mechanisms that underlie IS and COVID-19 are not well understood. Therefore, we implemented transcriptomic analysis from eight GEO datasets consist of 1191 samples to detect common pathways and molecular biomarkers in IS and COVID-19 that help understand the linkage between them. Differentially expressed genes (DEGs) were detected for IS and COVID-19 separately for finding shared mechanisms and we found that immune-related pathways were outlined with statistical significance. JAK2, which was identified as a hub gene, was supposed to be a potential therapeutic gene targets during the immunological process of COVID-19 and IS. Besides, we found a decrease in the proportion of CD8+ T and T helper 2 cells in the peripheral circulation of both COVID and IS patients, and NCR3 expression was significantly correlated with this change. In conclusion, we demonstrated that transcriptomic analyses reported in this study could make a deeper understanding of the common mechanism and might be promising for effective therapeutic for IS and COVID-19.

An evolutionary approach to the discretization of gene expression profiles to predict the severity of COVID-19

In this work, we propose to use a state-of-the-art evolutionary algorithm to set the discretization thresholds for gene expression profiles, using feedback from a classifier in order to maximize the accuracy of the predictions based on the discretized gene expression levels, while at the same time minimizing the number of different profiles obtained, to ease the understanding of the expert. The methodology is applied to a dataset containing COVID-19 patients that developed either mild or severe symptoms. The results show that the evolutionary approach performs better than a traditional discretization based on statistical analysis, and that it does preserve the sense-making necessary for practitioners to trust the results. © 2022 Owner/Author.

Identifying Pathways and Networks Associated With the SARS-CoV-2 Cell Receptor ACE2 Based on Gene Expression Profiles in Normal and SARS-CoV-2-Infected Human Tissues.

Because ACE2 is a host cell receptor of the SARS-CoV-2, an investigation of ACE2 expression in normal and virus-infected human tissues is crucial for understanding the mechanism of SARS-CoV-2 infection. We identified pathways associated with ACE2 expression and gene co-expression networks of ACE2 in pan-tissue based on the gene expression profiles in normal human tissues. We found that the pathways significantly associated with ACE2 upregulation were mainly involved in immune, stromal signature, metabolism, cell growth and proliferation, and cancer and other diseases. The number of genes having a significant positive expression correlation with ACE2 in females far exceeded that in males. The estrogen receptors (ESR1 and ESR2) and androgen receptor (AR) genes had a significant positive expression correlation with ACE2. Meanwhile, the enrichment levels of immune cells were positively associated with the expression levels of ESR1 and ESR2, while they were inversely associated with the expression levels of AR in pan-tissue and multiple individual tissues. It suggests that females are likely to have a more robust immune defense system against SARS-CoV-2 than males. ACE2 was upregulated in SARS-CoV-2-infected tissues relative to normal tissues and in SARS-CoV-2-infected males relative to females, while its expression levels had no significant difference between healthy females and males. Numerous immune-related pathways were highly enriched in SARS-CoV-2-infected males relative to females. These data indicate that males are more susceptible and more likely to have an excessive immune response to SARS-CoV-2 infection than females. This study furnishes potentially cues explaining why females have better clinical outcomes of SARS-CoV-2 infections than males and warrant further investigation for understanding the mechanism of SARS-CoV-2 infection.

Computational analysis of TMPRSS2 expression in normal and SARS-CoV-2-infected human tissues.

The transmembrane serine protease 2 (TMPRSS2) is a key molecule for SARS-CoV-2 invading human host cells. To provide insights into SARS-CoV-2 infection of various human tissues and understand the potential mechanism of SARS-CoV-2 infection, we investigated TMPRSS2 expression in various normal human tissues and SARS-CoV-2-infected human tissues. Using publicly available datasets, we performed computational analyses of TMPRSS2 expression levels in 30 normal human tissues, and compared them between males and females and between younger (ages ≤ 49 years) and older (ages > 49 years) populations in these tissues. We found that TMPRSS2 expression levels were the highest in the prostate, stomach, pancreas, lungs, small intestine, and salivary gland. The TMPRSS2 protein had relatively high expression levels in the parathyroid gland, stomach, small intestine, pancreas, kidneys, seminal vesicle, epididymis, and prostate. However, TMPRSS2 expression levels were not significantly different between females and males or between younger and older populations in these tissues. The pathways enriched in TMPRSS2-upregulated pan-tissue were mainly involved in immune, metabolism, cell growth and proliferation, stromal signatures, and cancer and other diseases. Many cytokine genes displayed positive expression correlations with TMPRSS2 in pan-tissue, including TNF-α, IL-1, IL-2, IL-4, IL-7, IL-8, IL-12, IL-18, IFN-α, MCP-1, G-CSF, and IP-10. We further analyzed TMPRSS2 expression levels in nasopharyngeal swabs from SARS-CoV-2-infected patients. TMPRSS2 expression levels showed no significant difference between males and females or between younger and older patients. However, they were significantly lower in SARS-CoV-2-infected than in healthy individuals and patients with other viral acute respiratory illnesses. Interestingly, TMPRSS2 expression levels were positively correlated with the enrichment levels of four immune signatures (B cells, CD8+ T cells, NK cells, and interferon response) in SARS-CoV-2-infected patients but likely to be negatively correlated with them in the normal lung tissue. Our data may provide insights into the mechanism of SARS-CoV-2 infection.

Comparisons of the immunological landscape between COVID-19, influenza, and respiratory syncytial virus patients by clustering analysis.

BACKGROUND: COVID-19 has stronger infectivity and a higher risk for severity than most other contagious respiratory illnesses. The mechanisms underlying this difference remain unclear. METHODS: We compared the immunological landscape between COVID-19 and two other contagious respiratory illnesses (influenza and respiratory syncytial virus (RSV)) by clustering analysis of the three diseases based on 27 immune signatures' scores. RESULTS: We identified three immune subtypes: Immunity-H, Immunity-M, and Immunity-L, which displayed high, medium, and low immune signatures, respectively. We found 20%, 35.5%, and 44.5% of COVID-19 cases included in Immunity-H, Immunity-M, and Immunity-L, respectively; all influenza cases were included in Immunity-H; 66.7% and 33.3% of RSV cases belonged to Immunity-H and Immunity-L, respectively. These data indicate that most COVID-19 patients have weaker immune signatures than influenza and RSV patients, as evidenced by 22 of the 27 immune signatures having lower enrichment scores in COVID-19 than in influenza and/or RSV. The Immunity-M COVID-19 patients had the highest expression levels of ACE2 and IL-6 and lowest viral loads and were the youngest. In contrast, the Immunity-H COVID-19 patients had the lowest expression levels of ACE2 and IL-6 and highest viral loads and were the oldest. Most immune signatures had lower enrichment levels in the intensive care unit (ICU) than in non-ICU patients. Gene ontology analysis showed that the innate and adaptive immune responses were significantly downregulated in COVID-19 versus healthy individuals. CONCLUSIONS: Compared to influenza and RSV, COVID-19 displayed significantly different immunological profiles. Elevated immune signatures are associated with better prognosis in COVID-19 patients.

Identification of COVID-19 subtypes based on immunogenomic profiling.

Although previous studies have shown that the host immune response is crucial in determining clinical outcomes in COVID-19 patients, the association between host immune signatures and COVID-19 patient outcomes remains unclear. Based on the enrichment levels of 11 immune signatures (eight immune-inciting and three immune-inhibiting signatures) in leukocytes of 100 COVID-19 patients, we identified three COVID-19 subtypes: Im-C1, Im-C2, and Im-C3, by clustering analysis. Im-C1 had the lowest immune-inciting signatures and high immune-inhibiting signatures. Im-C2 had medium immune-inciting signatures and high immune-inhibiting signatures. Im-C3 had the highest immune-inciting signatures while the lowest immune-inhibiting signatures. Im-C3 and Im-C1 displayed the best and worst clinical outcomes, respectively, suggesting that antiviral immune responses alleviated the severity of COVID-19 patients. We further demonstrated that the adaptive immune response had a stronger impact on COVID-19 outcomes than the innate immune response. The patients in Im-C3 were younger than those in Im-C1, indicating that younger persons have stronger antiviral immune responses than older persons. Nevertheless, we did not observe a significant association between sex and immune responses in COVID-19 patients. In addition, we found that the type II IFN response signature was an adverse prognostic factor for COVID-19. Our identification of COVID-19 immune subtypes has potential clinical implications for the management of COVID-19 patients.