Identification of Human miRNA Biomarkers Targeting the SARS-CoV-2 Genome.

SARS-CoV-2 poses a great challenge toward mankind, majorly due to its evolution and frequently occurring variants. On the other hand, in human hosts, microRNA (miRNA) plays a vital role in replication and propagation during a viral infection and can control the biological processes. This may be essential for the progression of viral infection. Moreover, human miRNAs can play a therapeutic role in treatment of different viral diseases by binding to the target sites of the virus genome, thereby hindering the essential functioning of the virus. Motivated by this fact, we have hypothesized a new approach in order to identify human miRNAs that can target the mRNA (genome) of SARS-CoV-2 to degrade their protein synthesis. In this regard, the multiple sequence alignment technique Clustal Omega is used to align a complement of 2656 human miRNAs with the SARS-CoV-2 reference genome (mRNA). Thereafter, ranking of these aligned human miRNAs is performed with the help of a new scoring function that takes into account the (a) total number of nucleotide matches between the human miRNA and the SARS-CoV-2 genome, (b) number of consecutive nucleotide matches between the human miRNA and the SARS-CoV-2 genome, (c) number of nucleotide mismatches between the human miRNA and the SARS-CoV-2 genome, and (d) the difference in length before and after alignment of the human miRNA. As a result, from the 2656 ranked miRNAs, the top 20 human miRNAs are reported, which are targeting different coding and non-coding regions of the SARS-CoV-2 genome. Moreover, molecular docking of such human miRNAs with virus mRNA is performed to verify the efficacy of the interactions. Furthermore, 4 miRNAs out of the top 20 miRNAs are identified to have the seed region. In order to inhibit the virus, the key human targets of the seed regions may be targeted. Repurposable drugs like carfilzomib, bortezomib, hydralazine, and paclitaxel are identified for such purpose.

Unveiling the Biomarkers of Cancer and COVID-19 and Their Regulations in Different Organs by Integrating RNA-Seq Expression and Protein-Protein Interactions.

Cancer and COVID-19 have killed millions of people worldwide. COVID-19 is even more dangerous to people with comorbidities such as cancer. Thus, it is imperative to identify the key human genes or biomarkers that can be targeted to develop novel prognosis and therapeutic strategies. The transcriptomic data provided by the next-generation sequencing technique makes this identification very convenient. Hence, mRNA (messenger ribonucleic acid) expression data of 2265 cancer and 282 normal patients were considered, while for COVID-19 assessment, 784 and 425 COVID-19 and normal patients were taken, respectively. Initially, volcano plots were used to identify the up- and down-regulated genes for both cancer and COVID-19. Thereafter, protein-protein interaction (PPI) networks were prepared by combining all the up- and down-regulated genes for each of cancer and COVID-19. Subsequently, such networks were analyzed to identify the top 10 genes with the highest degree of connection to provide the biomarkers. Interestingly, these genes were all up-regulated for cancer, while they were down-regulated for COVID-19. This study had also identified common genes between cancer and COVID-19, all of which were up-regulated in both the diseases. This analysis revealed that FN1 was highly up-regulated in different organs for cancer, while EEF2 was dysregulated in most organs affected by COVID-19. Then, functional enrichment analysis was performed to identify significant biological processes. Finally, the drugs for cancer and COVID-19 biomarkers and the common genes between them were identified using the Enrichr online web tool. These drugs include lucanthone, etoposide, and methotrexate, targeting the biomarkers for cancer, while paclitaxel is an important drug for COVID-19.

ML-DTD: Machine Learning-Based Drug Target Discovery for the Potential Treatment of COVID-19.

Recent research has highlighted that a large section of druggable protein targets in the Human interactome remains unexplored for various diseases. It might lead to the drug repurposing study and help in the in-silico prediction of new drug-human protein target interactions. The same applies to the current pandemic of COVID-19 disease in global health issues. It is highly desirable to identify potential human drug targets for COVID-19 using a machine learning approach since it saves time and labor compared to traditional experimental methods. Structure-based drug discovery where druggability is determined by molecular docking is only appropriate for the protein whose three-dimensional structures are available. With machine learning algorithms, differentiating relevant features for predicting targets and non-targets can be used for the proteins whose 3-D structures are unavailable. In this research, a Machine Learning-based Drug Target Discovery (ML-DTD) approach is proposed where a machine learning model is initially built up and tested on the curated dataset consisting of COVID-19 human drug targets and non-targets formed by using the Therapeutic Target Database (TTD) and human interactome using several classifiers like XGBBoost Classifier, AdaBoost Classifier, Logistic Regression, Support Vector Classification, Decision Tree Classifier, Random Forest Classifier, Naive Bayes Classifier, and K-Nearest Neighbour Classifier (KNN). In this method, protein features include Gene Set Enrichment Analysis (GSEA) ranking, properties derived from the protein sequence, and encoded protein network centrality-based measures. Among all these, XGBBoost, KNN, and Random Forest models are satisfactory and consistent. This model is further used to predict novel COVID-19 human drug targets, which are further validated by target pathway analysis, the emergence of allied repurposed drugs, and their subsequent docking study.

Online Predictor Using Machine Learning to Predict Novel Coronavirus and Other Pathogenic Viruses

The problem of virus classification is always a subject of concern for virology or epidemiology over the decades. In this regard, a machine learning technique can be used to predict the novel coronavirus by considering its sequence. Thus, we are proposing a machine learning-based novel coronavirus prediction technique, called COVID-Predictor, where 1000 sequences of SARS-CoV-1, MERS-CoV, SARS-CoV-2, and other viruses are used to train a Naive Bayes classifier so that it can predict any unknown sequences of these viruses. The model has been validated using 10-fold cross-validation in comparison with other machine learning techniques. The results show the superiority of our predictor by achieving an average 99.7% accuracy on an unseen validation set of viruses. The same pre-trained model has been used to design a web-based application where sequences of unknown viruses can be uploaded to predict the novel coronavirus.

Drug repurposing for identification of potential spike inhibitors for SARS-CoV-2 using molecular docking and molecular dynamics simulations.

For the last two years, the COVID-19 pandemic has continued to bring consternation on most of the world. According to recent WHO estimates, there have been more than 5.6 million deaths worldwide. The virus continues to evolve all over the world, thus requiring both vigilance and the necessity to find and develop a variety of therapeutic treatments, including the identification of specific antiviral drugs. Multiple studies have confirmed that SARS-CoV-2 utilizes its membrane-bound spike protein to recognize human angiotensin-converting enzyme 2 (ACE2). Thus, preventing spike-ACE2 interactions is a potentially viable strategy for COVID-19 treatment as it would block the virus from binding and entering into a host cell. This work aims to identify potential drugs using an in silico approach. Molecular docking was carried out on both approved drugs and substances previously tested in vivo. This step was followed by a more detailed analysis of selected ligands by molecular dynamics simulations to identify the best molecules that thwart the ability of the virus to interact with the ACE2 receptor. Because the SARS-CoV-2 virus evolves rapidly due to a plethora of immunocompromised hosts, the compounds were tested against five different known lineages. As a result, we could identify substances that work well on individual lineages and those showing broader efficacy. The most promising candidates among the currently used drugs were zafirlukast and simeprevir with an average binding affinity of -22 kcal/mol for spike proteins originating from various lineages. The first compound is a leukotriene receptor antagonist that is used to treat asthma, while the latter is a protease inhibitor used for hepatitis C treatment. From among the in vivo tested substances that concurrently exhibit promising free energy of binding and ADME parameters (indicating a possible oral administration) we selected the compound BDBM50136234. In conclusion, these molecules are worth exploring further by in vitro and in vivo studies against SARS-CoV-2.

Genome-wide analysis of 10664 SARS-CoV-2 genomes to identify virus strains in 73 countries based on single nucleotide polymorphism.

Since the onslaught of SARS-CoV-2, the research community has been searching for a vaccine to fight against this virus. However, during this period, the virus has mutated to adapt to the different environmental conditions in the world and made the task of vaccine design more challenging. In this situation, the identification of virus strains is very much timely and important task. We have performed genome-wide analysis of 10664 SARS-CoV-2 genomes of 73 countries to identify and prepare a Single Nucleotide Polymorphism (SNP) dataset of SARS-CoV-2. Thereafter, with the use of this SNP data, the advantage of hierarchical clustering is taken care of in such a way so that Average Linkage and Complete Linkage with Jaccard and Hamming distance functions are applied separately in order to identify the virus strains as clusters present in the SNP data. In this regard, the consensus of both the clustering results are also considered while Silhouette index is used as a cluster validity index to measure the goodness of the clusters as well to determine the number of clusters or virus strains. As a result, we have identified five major clusters or virus strains present worldwide. Apart from quantitative measures, these clusters are also visualized using Visual Assessment of Tendency (VAT) plot. The evolution of these clusters are also shown. Furthermore, top 10 signature SNPs are identified in each cluster and the non-synonymous signature SNPs are visualised in the respective protein structures. Also, the sequence and structural homology-based prediction along with the protein structural stability of these non-synonymous signature SNPs are reported in order to judge the characteristics of the identified clusters. As a consequence, T85I, Q57H and R203M in NSP2, ORF3a and Nucleocapsid respectively are found to be responsible for Cluster 1 as they are damaging and unstable non-synonymous signature SNPs. Similarly, F506L and S507C in Exon are responsible for both Clusters 3 and 4 while Clusters 2 and 5 do not exhibit such behaviour due to the absence of any non-synonymous signature SNPs. In addition to all these, the code, SNP dataset, 10664 labelled SARS-CoV-2 strains and additional results as supplementary are provided through our website for further use.

COVID-DeepPredictor: Recurrent Neural Network to Predict SARS-CoV-2 and Other Pathogenic Viruses.

The COVID-19 disease for Novel coronavirus (SARS-CoV-2) has turned out to be a global pandemic. The high transmission rate of this pathogenic virus demands an early prediction and proper identification for the subsequent treatment. However, polymorphic nature of this virus allows it to adapt and sustain in different kinds of environment which makes it difficult to predict. On the other hand, there are other pathogens like SARS-CoV-1, MERS-CoV, Ebola, Dengue, and Influenza as well, so that a predictor is highly required to distinguish them with the use of their genomic information. To mitigate this problem, in this work COVID-DeepPredictor is proposed on the framework of deep learning to identify an unknown sequence of these pathogens. COVID-DeepPredictor uses Long Short Term Memory as Recurrent Neural Network for the underlying prediction with an alignment-free technique. In this regard, k-mer technique is applied to create Bag-of-Descriptors (BoDs) in order to generate Bag-of-Unique-Descriptors (BoUDs) as vocabulary and subsequently embedded representation is prepared for the given virus sequences. This predictor is not only validated for the dataset using K -fold cross-validation but also for unseen test datasets of SARS-CoV-2 sequences and sequences from other viruses as well. To verify the efficacy of COVID-DeepPredictor, it has been compared with other state-of-the-art prediction techniques based on Linear Discriminant Analysis, Random Forests, and Gradient Boosting Method. COVID-DeepPredictor achieves 100% prediction accuracy on validation dataset while on test datasets, the accuracy ranges from 99.51 to 99.94%. It shows superior results over other prediction techniques as well. In addition to this, accuracy and runtime of COVID-DeepPredictor are considered simultaneously to determine the value of k in k-mer, a comparative study among k values in k-mer, Bag-of-Descriptors (BoDs), and Bag-of-Unique-Descriptors (BoUDs) and a comparison between COVID-DeepPredictor and Nucleotide BLAST have also been performed. The code, training, and test datasets used for COVID-DeepPredictor are available at http://www.nitttrkol.ac.in/indrajit/projects/COVID-DeepPredictor/.