A crowdsourced analysis to identify ab initio molecular signatures predictive of susceptibility to viral infection

Clinical microbiology and artificial intelligence: Different applications, challenges, and future prospects

Conventional clinical microbiological techniques are enhanced by the introduction of artificial intelligence (AI). Comprehensive data processing and analysis enabled the development of curated datasets that has been effectively used in training different AI algorithms. Recently, a number of machine learning (ML) and deep learning (DL) algorithms are developed and evaluated using diverse microbiological datasets. These datasets included spectral analysis (Raman and MALDI-TOF spectroscopy), microscopic images (Gram and acid fast stains), and genomic and protein sequences (whole genome sequencing (WGS) and protein data banks (PDBs)). The primary objective of these algorithms is to minimize the time, effort, and expenses linked to conventional analytical methods. Furthermore, AI algorithms are incorporated with quantitative structure-activity relationship (QSAR) models to predict novel antimicrobial agents that address the continuing surge of antimicrobial resistance. During the COVID-19 pandemic, AI algorithms played a crucial role in vaccine developments and the discovery of new antiviral agents, and introduced potential drug candidates via drug repurposing. However, despite their significant benefits, the implementation of AI encounters various challenges, including ethical considerations, the potential for bias, and errors related to data training. This review seeks to provide an overview of the most recent applications of artificial intelligence in clinical microbiology, with the intention of educating a wider audience of clinical practitioners regarding the current uses of machine learning algorithms and encouraging their implementation. Furthermore, it will discuss the challenges related to the incorporation of AI into clinical microbiology laboratories and examine future opportunities for AI within the realm of infectious disease epidemiology.

Predictive signatures of immune response to vaccination and implications of the immune setpoint remodeling

In 2020, I featured two articles in the "mSphere of Influence" commentary series that had profound implications for the field of immunology and helped shape my research perspective. These articles were "Global Analyses of Human Immune Variation Reveal Baseline Predictors of Postvaccination Responses" by Tsang et al. (Cell 157:499-513, 2014, https://doi.org/10.1016/j.cell.2014.03.031) and "A crowdsourced analysis to identify ab initio molecular signatures predictive of susceptibility to viral infection" by Fourati et al. (Nat Commun 9:4418, 2018, https://doi.org/10.1038/s41467-018-06735-8). From these topics, the identification of signatures predictive of immune responses to vaccination has greatly advanced and pivoted our understanding of how the immune state at the time of vaccination predicts (and potentially determines) vaccination outcomes. While most of this work has been done using influenza vaccination as a model, pan-vaccine signatures have been also identified. The key implications are their potential use to predict who will respond to vaccinations and to inform strategies for fine-tuning the immune setpoint to enhance immune responses. In addition, investigations in this area led us to understand that immune perturbations, such as acute infections and vaccinations, can remodel the baseline immune state and alter immune responses to future exposures, expanding this exciting field of research. These processes are likely epigenetically encoded, and some examples have already been identified and are discussed in this minireview. Therefore, further research is essential to gain a deeper understanding of how immune exposures modify the epigenome and transcriptome, influence the immune setpoint in response to vaccination, and define its exposure-specific characteristics.

Comparative analysis of machine learning approaches for predicting respiratory virus infection and symptom severity

Respiratory diseases are among the major health problems causing a burden on hospitals. Diagnosis of infection and rapid prediction of severity without time-consuming clinical tests could be beneficial in preventing the spread and progression of the disease, especially in countries where health systems remain incapable. Personalized medicine studies involving statistics and computer technologies could help to address this need. In addition to individual studies, competitions are also held such as Dialogue for Reverse Engineering Assessment and Methods (DREAM) challenge which is a community-driven organization with a mission to research biology, bioinformatics, and biomedicine. One of these competitions was the Respiratory Viral DREAM Challenge, which aimed to develop early predictive biomarkers for respiratory virus infections. These efforts are promising, however, the prediction performance of the computational methods developed for detecting respiratory diseases still has room for improvement. In this study, we focused on improving the performance of predicting the infection and symptom severity of individuals infected with various respiratory viruses using gene expression data collected before and after exposure. The publicly available gene expression dataset in the Gene Expression Omnibus, named GSE73072, containing samples exposed to four respiratory viruses (H1N1, H3N2, human rhinovirus (HRV), and respiratory syncytial virus (RSV)) was used as input data. Various preprocessing methods and machine learning algorithms were implemented and compared to achieve the best prediction performance. The experimental results showed that the proposed approaches obtained a prediction performance of 0.9746 area under the precision-recall curve (AUPRC) for infection (i.e., shedding) prediction (SC-1), 0.9182 AUPRC for symptom class prediction (SC-2), and 0.6733 Pearson correlation for symptom score prediction (SC-3) by outperforming the best leaderboard scores of Respiratory Viral DREAM Challenge (a 4.48% improvement for SC-1, a 13.68% improvement for SC-2, and a 13.98% improvement for SC-3). Additionally, over-representation analysis (ORA), which is a statistical method for objectively determining whether certain genes are more prevalent in pre-defined sets such as pathways, was applied using the most significant genes selected by feature selection methods. The results show that pathways associated with the 'adaptive immune system' and 'immune disease' are strongly linked to pre-infection and symptom development. These findings contribute to our knowledge about predicting respiratory infections and are expected to facilitate the development of future studies that concentrate on predicting not only infections but also the associated symptoms.

Using three statistical methods to analyze the association between aldehyde exposure and markers of inflammation and oxidative stress

BACKGROUND

Exposure to aldehydes has been linked to adverse health outcomes such as inflammation and oxidative stress, but research on the effects of these compounds is limited. This study is aimed at assessing the association between aldehyde exposure and markers of inflammation and oxidative stress.

METHODS

The study used data from the NHANES 2013-2014 survey (n = 766) and employed multivariate linear models to investigate the relationship between aldehyde compounds and various markers of inflammation (alkaline phosphatase (ALP) level, absolute neutrophil count (ANC), and lymphocyte count) and oxidative stress (bilirubin, albumin, and iron levels) while controlling for other relevant factors. In addition to generalized linear regression, weighted quantile sum (WQS) and Bayesian kernel machine regression (BKMR) analyses were applied to examine the single or overall effect of aldehyde compounds on the outcomes.

RESULTS

In the multivariate linear regression model, each 1 standard deviation (SD) change in propanaldehyde and butyraldehyde was significantly associated with increases in serum iron levels (beta and 95% confidence interval, 3.25 (0.24, 6.27) and 8.40 (0.97, 15.83), respectively) and the lymphocyte count (0.10 (0.04, 0.16) and 0.18 (0.03, 0.34), respectively). In the WQS regression model, a significant association was discovered between the WQS index and both the albumin and iron levels. Furthermore, the results of the BKMR analysis showed that the overall impact of aldehyde compounds was significantly and positively correlated with the lymphocyte count, as well as the levels of albumin and iron, suggesting that these compounds may contribute to increased oxidative stress.

CONCLUSIONS

This study reveals the close association between single or overall aldehyde compounds and markers of chronic inflammation and oxidative stress, which has essential guiding value for exploring the impact of environmental pollutants on population health.

Susceptibility identification for seasonal influenza A/H3N2 based on baseline blood transcriptome

Introduction

Influenza susceptibility difference is a widely existing trait that has great practical significance for the accurate prevention and control of influenza.

Methods

Here, we focused on the human susceptibility to the seasonal influenza A/H3N2 of healthy adults at baseline level. Whole blood expression data for influenza A/H3N2 susceptibility from GEO were collected firstly (30 symptomatic and 19 asymptomatic). Then to explore the differences at baseline, a suite of systems biology approaches - the differential expression analysis, co-expression network analysis, and immune cell frequencies analysis were utilized.

Results

We found the baseline condition, especially immune condition between symptomatic and asymptomatic, was different. Co-expression module that is positively related to asymptomatic is also related to immune cell type of naïve B cell. Function enrichment analysis showed significantly correlation with "B cell receptor signaling pathway", "immune response-activating cell surface receptor signaling pathway" and so on. Also, modules that are positively related to symptomatic are also correlated to immune cell type of neutrophils, with function enrichment analysis showing significantly correlations with "response to bacterium", "inflammatory response", "cAMP-dependent protein kinase complex" and so on. Responses of symptomatic and asymptomatic hosts after virus exposure show differences on resisting the virus, with more effective frontline defense for asymptomatic hosts. A prediction model was also built based on only baseline transcription information to differentiate symptomatic and asymptomatic population with accuracy of 0.79.

Discussion

The results not only improve our understanding of the immune system and influenza susceptibility, but also provide a new direction for precise and targeted prevention and therapy of influenza.

DeepFlu: a deep learning approach for forecasting symptomatic influenza A infection based on pre-exposure gene expression

BACKGROUND AND OBJECTIVE

Not everyone gets sick after an exposure to influenza A viruses (IAV). Although KLRD1 has been identified as a potential biomarker for influenza susceptibility, it remains unclear whether forecasting symptomatic flu infection based on pre-exposure host gene expression might be possible.

METHOD

To examine this hypothesis, we developed DeepFlu using the state-of-the-art deep learning approach on the human gene expression data infected with IAV subtype H1N1 or H3N2 viruses to forecast who would catch the flu prior to an exposure to IAV.

RESULTS

The results indicated that such forecast is possible and, in other words, gene expression could reflect the strength of host immunity. In the leave-one-person-out cross-validation, DeepFlu based on deep neural network outperformed the models using convolutional neural network, random forest, or support vector machine, achieving 70.0% accuracy, 0.787 AUROC, and 0.758 AUPR for H1N1 and 73.8% accuracy, 0.847 AUROC, and 0.901 AUPR for H3N2. In the external validation, DeepFlu also reached 71.4% accuracy, 0.700 AUROC, and 0.723 AUPR for H1N1 and 73.5% accuracy, 0.732 AUROC, and 0.749 AUPR for H3N2, surpassing the KLRD1 biomarker. In addition, DeepFlu which was trained only by pre-exposure data worked the best than by other time spans and mixed training data of H1N1 and H3N2 did not necessarily enhance prediction. DeepFlu is available at https://github.com/ntou-compbio/DeepFlu.

CONCLUSIONS

DeepFlu is a prognostic tool that can moderately recognize individuals susceptible to the flu and may help prevent the spread of IAV.

Causal Effects of Genetically Predicted Iron Status on Sepsis: A Two-Sample Bidirectional Mendelian Randomization Study

Background/Aim: Several observational studies showed a significant association between elevated iron status biomarkers levels and sepsis with the unclear direction of causality. A two-sample bidirectional mendelian randomization (MR) study was designed to identify the causal direction between seven iron status traits and sepsis. Methods: Seven iron status traits were studied, including serum iron, ferritin, transferrin saturation, transferrin, hemoglobin, erythrocyte count, and reticulocyte count. MR analysis was first performed to estimate the causal effect of iron status on the risk of sepsis and then performed in the opposite direction. The multiplicative random-effects and fixed-effects inverse-variance weighted, weighted median-based method and MR-Egger were applied. MR-Egger regression, MR pleiotropy residual sum and outlier (MR-PRESSO), and Cochran's Q statistic methods were used to assess heterogeneity and pleiotropy. Results: Genetically predicted high levels of serum iron (OR = 1.21, 95%CI = 1.13-1.29, p = 3.16 × 10^-4), ferritin (OR = 1.32, 95%CI = 1.07-1.62, p =0.009) and transferrin saturation (OR = 1.14, 95%CI = 1.06-1.23, p = 5.43 × 10^-4) were associated with an increased risk of sepsis. No significant causal relationships between sepsis and other four iron status biomarkers were observed. Conclusions: This present bidirectional MR analysis suggested the causal association of the high iron status with sepsis susceptibility, while the reverse causality hypothesis did not hold. The levels of transferrin, hemoglobin, erythrocytes, and reticulocytes were not significantly associated with sepsis. Further studies will be required to confirm the potential clinical value of such a prevention and treatment strategy.

Early prognosis of respiratory virus shedding in humans

This paper addresses the development of predictive models for distinguishing pre-symptomatic infections from uninfected individuals. Our machine learning experiments are conducted on publicly available challenge studies that collected whole-blood transcriptomics data from individuals infected with HRV, RSV, H1N1, and H3N2. We address the problem of identifying discriminatory biomarkers between controls and eventual shedders in the first 32 h post-infection. Our exploratory analysis shows that the most discriminatory biomarkers exhibit a strong dependence on time over the course of the human response to infection. We visualize the feature sets to provide evidence of the rapid evolution of the gene expression profiles. To quantify this observation, we partition the data in the first 32 h into four equal time windows of 8 h each and identify all discriminatory biomarkers using sparsity-promoting classifiers and Iterated Feature Removal. We then perform a comparative machine learning classification analysis using linear support vector machines, artificial neural networks and Centroid-Encoder. We present a range of experiments on different groupings of the diseases to demonstrate the robustness of the resulting models.

mSphere of Influence: Predicting Immune Responses and Susceptibility to Influenza Virus-May the Data Be with You

Irene Ramos works in the field of immunology to viral infections. In this mSphere of Influence article, she reflects on how “Global analyses of human immune variation reveal baseline predictors of postvaccination responses” by Tsang et al. (Cell 157:499–513, 2014, https://doi.org/10.1016/j.cell.2014.03.031) and “A crowdsourced analysis to identify ab initio molecular signatures predictive of susceptibility to viral infection” by Fourati et al.

Irene Ramos works in the field of immunology to viral infections. In this mSphere of Influence article, she reflects on how “Global analyses of human immune variation reveal baseline predictors of postvaccination responses” by Tsang et al. (Cell 157:499–513, 2014, https://doi.org/10.1016/j.cell.2014.03.031) and “A crowdsourced analysis to identify ab initio molecular signatures predictive of susceptibility to viral infection” by Fourati et al. (Nat Commun 9:4418, 2018, https://doi.org/10.1038/s41467-018-06735-8) made an impact on her by highlighting the importance of data science methods to understand virus-host interactions.

Host-Based Diagnostics for Acute Respiratory Infections

PURPOSE

The inappropriate use of antimicrobials, especially in acute respiratory infections (ARIs), is largely driven by difficulty distinguishing bacterial, viral, and noninfectious etiologies of illness. A new frontier in infectious disease diagnostics looks to the host response for disease classification. This article examines how host response-based diagnostics for ARIs are being used in clinical practice, as well as new developments in the research pipeline.

METHODS

A limited search was conducted of the relevant literature, with emphasis placed on literature published in the last 5 years (2014-2019).

FINDINGS

Advances are being made in all areas of host response-based diagnostics for ARIs. Specifically, there has been significant progress made in single protein biomarkers, as well as in various "omics" fields (including proteomics, metabolomics, and transcriptomics) and wearable technologies. There are many potential applications of a host response-based approach; a few key examples include the ability to discriminate bacterial and viral disease, presymptomatic diagnosis of infection, and pathogen-specific host response diagnostics, including modeling disease progression.

IMPLICATIONS

As biomarker measurement technologies continue to improve, host response-based diagnostics will increasingly be translated to clinically available platforms that can generate a holistic characterization of an individual's health. This knowledge, in the hands of both patient and provider, can improve care for the individual patient and help fight rising rates of antibiotic resistance.

Machine Learning SNP Based Prediction for Precision Medicine

In the past decade, precision genomics based medicine has emerged to provide tailored and effective healthcare for patients depending upon their genetic features. Genome Wide Association Studies have also identified population based risk genetic variants for common and complex diseases. In order to meet the full promise of precision medicine, research is attempting to leverage our increasing genomic understanding and further develop personalized medical healthcare through ever more accurate disease risk prediction models. Polygenic risk scoring and machine learning are two primary approaches for disease risk prediction. Despite recent improvements, the results of polygenic risk scoring remain limited due to the approaches that are currently used. By contrast, machine learning algorithms have increased predictive abilities for complex disease risk. This increase in predictive abilities results from the ability of machine learning algorithms to handle multi-dimensional data. Here, we provide an overview of polygenic risk scoring and machine learning in complex disease risk prediction. We highlight recent machine learning application developments and describe how machine learning approaches can lead to improved complex disease prediction, which will help to incorporate genetic features into future personalized healthcare. Finally, we discuss how the future application of machine learning prediction models might help manage complex disease by providing tissue-specific targets for customized, preventive interventions.