Ranking of cell clusters in a single-cell RNA-sequencing analysis framework using prior knowledge.

Prioritization or ranking of different cell types in a single-cell RNA sequencing (scRNA-seq) framework can be performed in a variety of ways, some of these include: i) obtaining an indication of the proportion of cell types between the different conditions under study, ii) counting the number of differentially expressed genes (DEGs) between cell types and conditions in the experiment or, iii) prioritizing cell types based on prior knowledge about the conditions under study (i.e., a specific disease). These methods have drawbacks and limitations thus novel methods for improving cell ranking are required. Here we present a novel methodology that exploits prior knowledge in combination with expert-user information to accentuate cell types from a scRNA-seq analysis that yield the most biologically meaningful results with respect to a disease under study. Our methodology allows for ranking and prioritization of cell types based on how well their expression profiles relate to the molecular mechanisms and drugs associated with a disease. Molecular mechanisms, as well as drugs, are incorporated as prior knowledge in a standardized, structured manner. Cell types are then ranked/prioritized based on how well results from data-driven analysis of scRNA-seq data match the predefined prior knowledge. In additional cell-cell communication perturbations between disease and control networks are used to further prioritize/rank cell types. Our methodology has substantial advantages to more traditional cell ranking techniques and provides an informative complementary methodology that utilizes prior knowledge in a rapid and automated manner, that has previously not been attempted by other studies. The current methodology is also implemented as an R package entitled Single Cell Ranking Analysis Toolkit (scRANK) and is available for download and installation via GitHub (https://github.com/aoulas/scRANK).

Multi-omics data integration and network-based analysis drives a multiplex drug repurposing approach to a shortlist of candidate drugs against COVID-19.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic is undeniably the most severe global health emergency since the 1918 Influenza outbreak. Depending on its evolutionary trajectory, the virus is expected to establish itself as an endemic infectious respiratory disease exhibiting seasonal flare-ups. Therefore, despite the unprecedented rally to reach a vaccine that can offer widespread immunization, it is equally important to reach effective prevention and treatment regimens for coronavirus disease 2019 (COVID-19). Contributing to this effort, we have curated and analyzed multi-source and multi-omics publicly available data from patients, cell lines and databases in order to fuel a multiplex computational drug repurposing approach. We devised a network-based integration of multi-omic data to prioritize the most important genes related to COVID-19 and subsequently re-rank the identified candidate drugs. Our approach resulted in a highly informed integrated drug shortlist by combining structural diversity filtering along with experts' curation and drug-target mapping on the depicted molecular pathways. In addition to the recently proposed drugs that are already generating promising results such as dexamethasone and remdesivir, our list includes inhibitors of Src tyrosine kinase (bosutinib, dasatinib, cytarabine and saracatinib), which appear to be involved in multiple COVID-19 pathophysiological mechanisms. In addition, we highlight specific immunomodulators and anti-inflammatory drugs like dactolisib and methotrexate and inhibitors of histone deacetylase like hydroquinone and vorinostat with potential beneficial effects in their mechanisms of action. Overall, this multiplex drug repurposing approach, developed and utilized herein specifically for SARS-CoV-2, can offer a rapid mapping and drug prioritization against any pathogen-related disease.

Novel clinical, molecular and bioinformatics insights into the genetic background of autism.

BACKGROUND: Clinical classification of autistic patients based on current WHO criteria provides a valuable but simplified depiction of the true nature of the disorder. Our goal is to determine the biology of the disorder and the ASD-associated genes that lead to differences in the severity and variability of clinical features, which can enhance the ability to predict clinical outcomes. METHOD: Novel Whole Exome Sequencing data from children (n = 33) with ASD were collected along with extended cognitive and linguistic assessments. A machine learning methodology and a literature-based approach took into consideration known effects of genetic variation on the translated proteins, linking them with specific ASD clinical manifestations, namely non-verbal IQ, memory, attention and oral language deficits. RESULTS: Linear regression polygenic risk score results included the classification of severe and mild ASD samples with a 81.81% prediction accuracy. The literature-based approach revealed 14 genes present in all sub-phenotypes (independent of severity) and others which seem to impair individual ones, highlighting genetic profiles specific to mild and severe ASD, which concern non-verbal IQ, memory, attention and oral language skills. CONCLUSIONS: These genes can potentially contribute toward a diagnostic gene-set for determining ASD severity. However, due to the limited number of patients in this study, our classification approach is mostly centered on the prediction and verification of these genes and does not hold a diagnostic nature per se. Substantial further experimentation is required to validate their role as diagnostic markers. The use of these genes as input for functional analysis highlights important biological processes and bridges the gap between genotype and phenotype in ASD.

Circulating small RNA signatures differentiate accurately the subtypes of muscular dystrophies: small-RNA next-generation sequencing analytics and functional insights.

Muscular dystrophies are a group of rare and severe inherited disorders mainly affecting the muscle tissue. Duchene Muscular Dystrophy, Myotonic Dystrophy types 1 and 2, Limb Girdle Muscular Dystrophy and Facioscapulohumeral Muscular Dystrophy are some of the members of this family of disorders. In addition to the current diagnostic tools, there is an increasing interest for the development of novel non-invasive biomarkers for the diagnosis and monitoring of these diseases. miRNAs are small RNA molecules characterized by high stability in blood thus making them ideal biomarker candidates for various diseases. In this study, we present the first genome-wide next-generation small RNA sequencing in serum samples of five different types of muscular dystrophy patients and healthy individuals. We identified many small RNAs including miRNAs, lncRNAs, tRNAs, snoRNAs and snRNAs, that differentially discriminate the muscular dystrophy patients from the healthy individuals. Further analysis of the identified miRNAs showed that some miRNAs can distinguish the muscular dystrophy patients from controls and other miRNAs are specific to the type of muscular dystrophy. Bioinformatics analysis of the target genes for the most significant miRNAs and the biological role of these genes revealed different pathways that the dysregulated miRNAs are involved in each type of muscular dystrophy investigated. In conclusion, this study shows unique signatures of small RNAs circulating in five types of muscular dystrophy patients and provides a useful resource for future studies for the development of miRNA biomarkers in muscular dystrophies and for their involvement in the pathogenesis of the disorders.

Systems Bioinformatics: increasing precision of computational diagnostics and therapeutics through network-based approaches.

Systems Bioinformatics is a relatively new approach, which lies in the intersection of systems biology and classical bioinformatics. It focuses on integrating information across different levels using a bottom-up approach as in systems biology with a data-driven top-down approach as in bioinformatics. The advent of omics technologies has provided the stepping-stone for the emergence of Systems Bioinformatics. These technologies provide a spectrum of information ranging from genomics, transcriptomics and proteomics to epigenomics, pharmacogenomics, metagenomics and metabolomics. Systems Bioinformatics is the framework in which systems approaches are applied to such data, setting the level of resolution as well as the boundary of the system of interest and studying the emerging properties of the system as a whole rather than the sum of the properties derived from the system's individual components. A key approach in Systems Bioinformatics is the construction of multiple networks representing each level of the omics spectrum and their integration in a layered network that exchanges information within and between layers. Here, we provide evidence on how Systems Bioinformatics enhances computational therapeutics and diagnostics, hence paving the way to precision medicine. The aim of this review is to familiarize the reader with the emerging field of Systems Bioinformatics and to provide a comprehensive overview of its current state-of-the-art methods and technologies. Moreover, we provide examples of success stories and case studies that utilize such methods and tools to significantly advance research in the fields of systems biology and systems medicine.

PathWalks: identifying pathway communities using a disease-related map of integrated information.

MOTIVATION: Understanding the underlying biological mechanisms and respective interactions of a disease remains an elusive, time consuming and costly task. Computational methodologies that propose pathway/mechanism communities and reveal respective relationships can be of great value as they can help expedite the process of identifying how perturbations in a single pathway can affect other pathways. RESULTS: We present a random-walks-based methodology called PathWalks, where a walker crosses a pathway-to-pathway network under the guidance of a disease-related map. The latter is a gene network that we construct by integrating multi-source information regarding a specific disease. The most frequent trajectories highlight communities of pathways that are expected to be strongly related to the disease under study.We apply the PathWalks methodology on Alzheimer's disease and idiopathic pulmonary fibrosis and establish that it can highlight pathways that are also identified by other pathway analysis tools as well as are backed through bibliographic references. More importantly, PathWalks produces additional new pathways that are functionally connected with those already established, giving insight for further experimentation. AVAILABILITY AND IMPLEMENTATION: https://github.com/vagkaratzas/PathWalks. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PathwayConnector: finding complementary pathways to enhance functional analysis.

SUMMARY: PathwayConnector is a web-tool that facilitates the construction of complementary pathway-to-pathway networks and subnetworks of them, based on a reference pathway network derived from the rich information available either in KEGG or Reactome database for pathway mapping. Specifically, for a given set of pathways, PathwayConnector (i) finds all the direct connections between them, (ii) adds a minimum set of complementary pathways required to achieve connectivity between the pathways, leading to informative fully connected networks and (ii) provides a series of clustering methods for the further grouping of pathways in to sub-clusters. The proposed web-tool is a simple yet informative tool towards identifying connected groups of pathways that are significantly related to specific diseases. AVAILABILITY AND IMPLEMENTATION: http://bioinformatics.cing.ac.cy/PathwayConnector. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Positioning Europe for the EPITRANSCRIPTOMICS challenge.

The genetic alphabet consists of the four letters: C, A, G, and T in DNA and C,A,G, and U in RNA. Triplets of these four letters jointly encode 20 different amino acids out of which proteins of all organisms are built. This system is universal and is found in all kingdoms of life. However, bases in DNA and RNA can be chemically modified. In DNA, around 10 different modifications are known, and those have been studied intensively over the past 20 years. Scientific studies on DNA modifications and proteins that recognize them gave rise to the large field of epigenetic and epigenomic research. The outcome of this intense research field is the discovery that development, ageing, and stem-cell dependent regeneration but also several diseases including cancer are largely controlled by the epigenetic state of cells. Consequently, this research has already led to the first FDA approved drugs that exploit the gained knowledge to combat disease. In recent years, the ~150 modifications found in RNA have come to the focus of intense research. Here we provide a perspective on necessary and expected developments in the fast expanding area of RNA modifications, termed epitranscriptomics.

Metagenomic investigation of the geologically unique Hellenic Volcanic Arc reveals a distinctive ecosystem with unexpected physiology.

Hydrothermal vents represent a deep, hot, aphotic biosphere where chemosynthetic primary producers, fuelled by chemicals from Earth's subsurface, form the basis of life. In this study, we examined microbial mats from two distinct volcanic sites within the Hellenic Volcanic Arc (HVA). The HVA is geologically and ecologically unique, with reported emissions of CO2 -saturated fluids at temperatures up to 220°C and a notable absence of macrofauna. Metagenomic data reveals highly complex prokaryotic communities composed of chemolithoautotrophs, some methanotrophs, and to our surprise, heterotrophs capable of anaerobic degradation of aromatic hydrocarbons. Our data suggest that aromatic hydrocarbons may indeed be a significant source of carbon in these sites, and instigate additional research into the nature and origin of these compounds in the HVA. Novel physiology was assigned to several uncultured prokaryotic lineages; most notably, a SAR406 representative is attributed with a role in anaerobic hydrocarbon degradation. This dataset, the largest to date from submarine volcanic ecosystems, constitutes a significant resource of novel genes and pathways with potential biotechnological applications.

Impaired clearance of apoptotic cells leads to HMGB1 release in the bone marrow of patients with myelodysplastic syndromes and induces TLR4-mediated cytokine production.

Excessive pro-inflammatory cytokine production in the bone marrow has been associated with the pathogenesis of myelodysplastic syndromes. We herein investigated the involvement of toll-like receptors and their endogenous ligands in the induction/maintenance of the inflammatory process in the marrow of patients with myelodysplastic syndromes. We evaluated the expression of toll-like receptors in marrow monocytes of patients (n=27) and healthy controls (n=25) by flow-cytometry and also assessed the activation of the respective signaling using a real-time polymerase chain reaction-based array. We measured the high mobility group box-1 protein, a toll-like receptor-4 ligand, in marrow plasma and long-term bone marrow culture supernatants by an enzyme-linked immunosorbent assay and we performed cross-over experiments using marrow plasma from patients and controls in the presence/absence of a toll-like receptor-4 inhibitor to evaluate the pro-inflammatory cytokine production by chemiluminescence. We assessed the apoptotic cell clearance capacity of patients' macrophages using a fluorescence microscopy-based assay. We found over-expression of toll-like receptor-4 in patients' marrow monocytes compared to that in controls; this over-expression was associated with up-modulation of 53 genes related to the respective signaling. Incubation of patients' monocytes with autologous, but not with normal, marrow plasma resulted in over-production of pro-inflammatory cytokines, an effect that was abrogated by the toll-like receptor-4 inhibitor suggesting that the pro-inflammatory cytokine production in myelodysplastic syndromes is largely mediated through toll-like receptor-4. The levels of high mobility group box-1 protein were increased in patients' marrow plasma and culture supernatants compared to the levels in controls. Patients' macrophages displayed an impaired capacity to engulf apoptotic cells and this defect was associated with excessive release of high mobility group box-1 protein by dying cells. A primary apoptotic cell clearance defect of marrow macrophages in myelodysplastic syndromes may contribute to the induction/maintenance of the inflammatory process through aberrant release of molecules inducing toll-like receptor-4 such as high mobility group box-1 protein.

Bacterial Wars-a tool for the prediction of bacterial predominance based on network analysis measures.

Bacterial Wars (BW) is a network-based tool that applies a two-step pipeline to display information on the competition of bacterial species found in the same microbiome. It utilizes antimicrobial peptide (AMP) sequence similarities to obtain a relationship between species. The working hypothesis (putative AMP defense) is that friendly species share sequence similarity among the putative AMPs of their proteomes and are therefore immune to their AMPs. This may not happen in competing bacterial species with dissimilar putative AMPs. Similarities in the putative AMPs of bacterial proteomes may be thus used to predict predominance. The tool provides insights as to which bacterial species are more likely to 'die' in a competing environmental niche.

Exploring microbial functional biodiversity at the protein family level-From metagenomic sequence reads to annotated protein clusters.

Metagenomics has enabled accessing the genetic repertoire of natural microbial communities. Metagenome shotgun sequencing has become the method of choice for studying and classifying microorganisms from various environments. To this end, several methods have been developed to process and analyze the sequence data from raw reads to end-products such as predicted protein sequences or families. In this article, we provide a thorough review to simplify such processes and discuss the alternative methodologies that can be followed in order to explore biodiversity at the protein family level. We provide details for analysis tools and we comment on their scalability as well as their advantages and disadvantages. Finally, we report the available data repositories and recommend various approaches for protein family annotation related to phylogenetic distribution, structure prediction and metadata enrichment.

A new microRNA target prediction tool identifies a novel interaction of a putative miRNA with CCND2.

Computational methods for miRNA target prediction vary in the algorithm used; and while one can state opinions about the strengths or weaknesses of each particular algorithm, the fact of the matter is that they fall substantially short of capturing the full detail of physical, temporal and spatial requirements of miRNA::target-mRNA interactions. Here, we introduce a novel miRNA target prediction tool called Targetprofiler that utilizes a probabilistic learning algorithm in the form of a hidden Markov model trained on experimentally verified miRNA targets. Using a large scale protein downregulation data set we validate our method and compare its performance to existing tools. We find that Targetprofiler exhibits greater correlation between computational predictions and protein downregulation and predicts experimentally verified miRNA targets more accurately than three other tools. Concurrently, we use primer extension to identify the mature sequence of a novel miRNA gene recently identified within a cancer associated genomic region and use Targetprofiler to predict its potential targets. Experimental verification of the ability of this small RNA molecule to regulate the expression of CCND2, a gene with documented oncogenic activity, confirms its functional role as a miRNA. These findings highlight the competitive advantage of our tool and its efficacy in extracting biologically significant results.

Serum miRNAs as biomarkers for the rare types of muscular dystrophy.

Muscular dystrophies are a group of disorders that cause progressive muscle weakness. There is an increasing interest for the development of biomarkers for these disorders and specifically for Duchene Muscular Dystrophy. Limited research however, has been performed on the biomarkers' development for the most rare muscular dystrophies, like the Facioscapulohumeral Muscular Dystrophy, Limb-Girdle Muscular Dystrophy and Myotonic Dystrophy type 2. Here, we aimed to identify novel serum-based miRNA biomarkers for these rare muscular dystrophies, through high-throughput next-generation RNA sequencing. We identified many miRNAs that associate with muscular dystrophy patients compared to controls. Based on a series of selection criteria, the two best candidate miRNAs for each of these disorders were chosen and validated in a larger number of patients. Our results showed that miR-223-3p and miR-206 are promising serum-based biomarkers for Facioscapulohumeral Muscular Dystrophy type 1, miR-143-3p and miR-486-3p for Limb-Girdle Muscular Dystrophy type 2A whereas miR-363-3p and miR-25-3p associate with Myotonic Dystrophy type 2. Some of the identified miRNAs were significantly elevated in the serum of the patients compared to controls, whereas some others were lower. In conclusion, we provide new evidence that certain circulating miRNAs may be used as biomarkers for three types of rare muscular dystrophies.

Transcriptomic characterization of tissues from patients and subsequent pathway analyses reveal biological pathways that are implicated in spastic ataxia.

BACKGROUND: Spastic ataxias (SAs) encompass a group of rare and severe neurodegenerative diseases, characterized by an overlap between ataxia and spastic paraplegia clinical features. They have been associated with pathogenic variants in a number of genes, including GBA2. This gene codes for the non-lysososomal ß-glucosylceramidase, which is involved in sphingolipid metabolism through its catalytic role in the degradation of glucosylceramide. However, the mechanism by which GBA2 variants lead to the development of SA is still unclear. METHODS: In this work, we perform next-generation RNA-sequencing (RNA-seq), in an attempt to discover differentially expressed genes (DEGs) in lymphoblastoid, fibroblast cell lines and induced pluripotent stem cell-derived neurons derived from patients with SA, homozygous for the GBA2 c.1780G > C missense variant. We further exploit DEGs in pathway analyses in order to elucidate candidate molecular mechanisms that are implicated in the development of the GBA2 gene-associated SA. RESULTS: Our data reveal a total of 5217 genes with significantly altered expression between patient and control tested tissues. Furthermore, the most significant extracted pathways are presented and discussed for their possible role in the pathogenesis of the disease. Among them are the oxidative stress, neuroinflammation, sphingolipid signaling and metabolism, PI3K-Akt and MAPK signaling pathways. CONCLUSIONS: Overall, our work examines for the first time the transcriptome profiles of GBA2-associated SA patients and suggests pathways and pathway synergies that could possibly have a role in SA pathogenesis. Lastly, it provides a list of DEGs and pathways that could be further validated towards the discovery of disease biomarkers.

Prediction of novel microRNA genes in cancer-associated genomic regions--a combined computational and experimental approach.

The majority of existing computational tools rely on sequence homology and/or structural similarity to identify novel microRNA (miRNA) genes. Recently supervised algorithms are utilized to address this problem, taking into account sequence, structure and comparative genomics information. In most of these studies miRNA gene predictions are rarely supported by experimental evidence and prediction accuracy remains uncertain. In this work we present a new computational tool (SSCprofiler) utilizing a probabilistic method based on Profile Hidden Markov Models to predict novel miRNA precursors. Via the simultaneous integration of biological features such as sequence, structure and conservation, SSCprofiler achieves a performance accuracy of 88.95% sensitivity and 84.16% specificity on a large set of human miRNA genes. The trained classifier is used to identify novel miRNA gene candidates located within cancer-associated genomic regions and rank the resulting predictions using expression information from a full genome tiling array. Finally, four of the top scoring predictions are verified experimentally using northern blot analysis. Our work combines both analytical and experimental techniques to show that SSCprofiler is a highly accurate tool which can be used to identify novel miRNA gene candidates in the human genome. SSCprofiler is freely available as a web service at http://www.imbb.forth.gr/SSCprofiler.html.

Land use in urban areas impacts the composition of soil bacterial communities involved in nitrogen cycling. A case study from Lefkosia (Nicosia) Cyprus.

The different types of land-use and soil lithology in urban and peri-urban areas of modern cities compose a complex mosaic of soil ecosystems. It is largely unknown how these differences result in changes in bacterial community composition and structure as well as in functional guilds involved in N cycling. To investigate the bacterial composition and the proportion of denitrifiers in agricultural, forested, schoolyard and industrial areas, 24 samples were collected from urban and peri-urban sites of Lefkosia. Bacterial diversity and the proportion of denitrifiers were assessed by NGS and qPCR, respectively. Proteobacteria, Actinobacteria, Bacteriodetes, Chloroflexi, Acidobacteria and Planctomycetes were identified as the most dominant phyla across all sites, while agricultural sites exhibited the highest bacterial diversity. Heavy metals such as Co, Pb, V and Al were identified as key factors shaping bacterial composition in industrial and schoolyard sites, while the bacterial assemblages in agricultural and forested sites were associated with Ca. Variance partitioning analysis showed that 10.2% of the bacterial community variation was explained by land use management, 5.1% by chemical elements due to soil lithology, and 1.4% by sampling location. The proportion of denitrifiers varied with land use management. In industrial and schoolyard sites, the abundance of the nosZII bacterial community increased while nirK abundance declined. Our data showed that land use and lithology have a moderate impact on the bacterial assemblages in urban and peri-urban areas of Lefkosia. As the nosZII bacterial community is important to the N2O sink capacity of soils, it would be interesting to elucidate the factors contributing to the proliferation of the nosZII clade in these soils.

Putative Antimicrobial Peptides Within Bacterial Proteomes Affect Bacterial Predominance: A Network Analysis Perspective.

The predominance of bacterial taxa in the gut, was examined in view of the putative antimicrobial peptide sequences (AMPs) within their proteomes. The working assumption was that compatible bacteria would share homology and thus immunity to their putative AMPs, while competing taxa would have dissimilarities in their proteome-hidden AMPs. A network-based method ("Bacterial Wars") was developed to handle sequence similarities of predicted AMPs among UniProt-derived protein sequences from different bacterial taxa, while a resulting parameter ("Die" score) suggested which taxa would prevail in a defined microbiome. T he working hypothesis was examined by correlating the calculated Die scores, to the abundance of bacterial taxa from gut microbiomes from different states of health and disease. Eleven publicly available 16S rRNA datasets and a dataset from a full shotgun metagenomics served for the analysis. The overall conclusion was that AMPs encrypted within bacterial proteomes affected the predominance of bacterial taxa in chemospheres.

In depth analysis of Cyprus-specific mutations of SARS-CoV-2 strains using computational approaches.

BACKGROUND: This study aims to characterize SARS-CoV-2 mutations which are primarily prevalent in the Cypriot population. Moreover, using computational approaches, we assess whether these mutations are associated with changes in viral virulence. METHODS: We utilize genetic data from 144 sequences of SARS-CoV-2 strains from the Cypriot population obtained between March 2020 and January 2021, as well as all data available from GISAID. We combine this with countries' regional information, such as deaths and cases per million, as well as COVID-19-related public health austerity measure response times. Initial indications of selective advantage of Cyprus-specific mutations are obtained by mutation tracking analysis. This entails calculating specific mutation frequencies within the Cypriot population and comparing these with their prevalence world-wide throughout the course of the pandemic. We further make use of linear regression models to extrapolate additional information that may be missed through standard statistical analysis. RESULTS: We report a single mutation found in the ORF1ab gene (nucleotide position 18,440) that appears to be significantly enriched within the Cypriot population. The amino acid change is denoted as S6059F, which maps to the SARS-CoV-2 NSP14 protein. We further analyse this mutation using regression models to investigate possible associations with increased deaths and cases per million. Moreover, protein structure prediction tools show that the mutation infers a conformational change to the protein that significantly alters its structure when compared to the reference protein. CONCLUSIONS: Investigating Cyprus-specific mutations for SARS-CoV-2 can lead to a better understanding of viral pathogenicity. Researching these mutations can generate potential links between viral-specific mutations and the unique genomics of the Cypriot population. This can not only lead to important findings from which to battle the pandemic on a national level, but also provide insights into viral virulence worldwide.

Generalized linear models provide a measure of virulence for specific mutations in SARS-CoV-2 strains.

This study aims to highlight SARS-COV-2 mutations which are associated with increased or decreased viral virulence. We utilize genetic data from all strains available from GISAID and countries' regional information, such as deaths and cases per million, as well as COVID-19-related public health austerity measure response times. Initial indications of selective advantage of specific mutations can be obtained from calculating their frequencies across viral strains. By applying modelling approaches, we provide additional information that is not evident from standard statistics or mutation frequencies alone. We therefore, propose a more precise way of selecting informative mutations. We highlight two interesting mutations found in genes N (P13L) and ORF3a (Q57H). The former appears to be significantly associated with decreased deaths and cases per million according to our models, while the latter shows an opposing association with decreased deaths and increased cases per million. Moreover, protein structure prediction tools show that the mutations infer conformational changes to the protein that significantly alter its structure when compared to the reference protein.