DeepHBV: a deep learning model to predict hepatitis B virus (HBV) integration sites

VIS Atlas: A Database of Virus Integration Sites in Human Genome from NGS Data to Explore Integration Patterns

Integration of oncogenic DNA viruses into the human genome is a key step in most virus-induced carcinogenesis. Here, we constructed a virus integration site (VIS) Atlas database, an extensive collection of integration breakpoints for three most prevalent oncoviruses, human papillomavirus, hepatitis B virus, and Epstein-Barr virus based on the next-generation sequencing (NGS) data, literature, and experimental data. There are 63,179 breakpoints and 47,411 junctional sequences with full annotations deposited in the VIS Atlas database, comprising 47 virus genotypes and 17 disease types. The VIS Atlas database provides (1) a genome browser for NGS breakpoint quality check, visualization of VISs, and the local genomic context; (2) a novel platform to discover integration patterns; and (3) a statistics interface for a comprehensive investigation of genotype-specific integration features. Data collected in the VIS Atlas aid to provide insights into virus pathogenic mechanisms and the development of novel antitumor drugs. The VIS Atlas database is available at https://www.vis-atlas.tech/.

Deep learning for detecting and elucidating human T-cell leukemia virus type 1 integration in the human genome

Human T-cell leukemia virus type 1 (HTLV-1), a retrovirus, is the causative agent for adult T cell leukemia/lymphoma and many other human diseases. Accurate and high throughput detection of HTLV-1 virus integration sites (VISs) across the host genomes plays a crucial role in the prevention and treatment of HTLV-1-associated diseases. Here, we developed DeepHTLV, the first deep learning framework for VIS prediction de novo from genome sequence, motif discovery, and cis-regulatory factor identification. We demonstrated the high accuracy of DeepHTLV with more efficient and interpretive feature representations. Decoding the informative features captured by DeepHTLV resulted in eight representative clusters with consensus motifs for potential HTLV-1 integration. Furthermore, DeepHTLV revealed interesting cis-regulatory elements in regulation of VISs that have significant association with the detected motifs. Literature evidence demonstrated nearly half (34) of the predicted transcription factors enriched with VISs were involved in HTLV-1-associated diseases. DeepHTLV is freely available at https://github.com/bsml320/DeepHTLV.

Artificial intelligence applied to omics data in liver diseases: Enhancing clinical predictions

Rapid development of biotechnology has led to the generation of vast amounts of multi-omics data, necessitating the advancement of bioinformatics and artificial intelligence to enable computational modeling to diagnose and predict clinical outcome. Both conventional machine learning and new deep learning algorithms screen existing data unbiasedly to uncover patterns and create models that can be valuable in informing clinical decisions. We summarized published literature on the use of AI models trained on omics datasets, with and without clinical data, to diagnose, risk-stratify, and predict survivability of patients with non-malignant liver diseases. A total of 20 different models were tested in selected studies. Generally, the addition of omics data to regular clinical parameters or individual biomarkers improved the AI model performance. For instance, using NAFLD fibrosis score to distinguish F0-F2 from F3-F4 fibrotic stages, the area under the curve (AUC) was 0.87. When integrating metabolomic data by a GMLVQ model, the AUC drastically improved to 0.99. The use of RF on multi-omics and clinical data in another study to predict progression of NAFLD to NASH resulted in an AUC of 0.84, compared to 0.82 when using clinical data only. A comparison of RF, SVM and kNN models on genomics data to classify immune tolerant phase in chronic hepatitis B resulted in AUC of 0.8793-0.8838 compared to 0.6759-0.7276 when using various serum biomarkers. Overall, the integration of omics was shown to improve prediction performance compared to models built only on clinical parameters, indicating a potential use for personalized medicine in clinical setting.

Hepatitis B and circadian rhythm of the liver

The circadian rhythm in humans is determined by the central clock located in the hypothalamus’s suprachiasmatic nucleus, and it synchronizes the peripheral clocks in other tissues. Circadian clock genes and clock-controlled genes exist in almost all cell types. They have an essential role in many physiological processes, including lipid metabolism in the liver, regulation of the immune system, and the severity of infections. In addition, circadian rhythm genes can stimulate the immune response of host cells to virus infection. Hepatitis B virus (HBV) infection is the leading cause of liver disease and liver cancer globally. HBV infection depends on the host cell, and hepatocyte circadian rhythm genes are associated with HBV replication, survival, and spread. The core circadian rhythm proteins, REV-ERB and brain and muscle ARNTL-like protein 1, have a crucial role in HBV replication in hepatocytes. In addition to influencing the virus’s life cycle, the circadian rhythm also affects the pharmacokinetics and efficacy of antiviral vaccines. Therefore, it is vital to apply antiviral therapy at the appropriate time of day to reduce toxicity and improve the effectiveness of antiviral treatment. For these reasons, understanding the role of the circadian rhythm in the regulation of HBV infection and host responses to the virus provides us with a new perspective of the interplay of the circadian rhythm and anti-HBV therapy. Therefore, this review emphasizes the importance of the circadian rhythm in HBV infection and the optimization of antiviral treatment based on the circadian rhythm-dependent immune response.

Decision Support System for Predicting Survivability of Hepatitis Patients

Background and Objective

Viral hepatitis is a major public health concern on a global scale. It predominantly affects the world's least developed countries. The most endemic regions are resource constrained, with a low human development index. Chronic hepatitis can lead to cirrhosis, liver failure, cancer and eventually death. Early diagnosis and treatment of hepatitis infection can help to reduce disease burden and transmission to those at risk of infection or reinfection. Screening is critical for meeting the WHO's 2030 targets. Consequently, automated systems for the reliable prediction of hepatitis illness. When applied to the prediction of hepatitis using imbalanced datasets from testing, machine learning (ML) classifiers and known methodologies for encoding categorical data have demonstrated a wide range of unexpected results. Early research also made use of an artificial neural network to identify features without first gaining a thorough understanding of the sequence data.

Methods

To help in accurate binary classification of diagnosis (survivability or mortality) in patients with severe hepatitis, this paper suggests a deep learning-based decision support system (DSS) that makes use of bidirectional long/short-term memory (BiLSTM). Balanced data was utilized to predict hepatitis using the BiLSTM model.

Results

In contrast to previous investigations, the trial results of this suggested model were encouraging: 95.08% accuracy, 94% precision, 93% recall, and a 93% F1-score.

Conclusions

In the field of hepatitis detection, the use of a BiLSTM model for classification is better than current methods by a significant margin in terms of improved accuracy.