key: cord-0297870-4bw6egae authors: Islam Bappy, Md. Nazmul; Robin, Tanjin Barketullah; Prome, Anindita Ash; Laskar, Fayeza Sadia; Roy, Anindita; Akter, Hafsa; Zinnah, Kazi Md. Ali title: Subtractive proteomics analysis to uncover the potent drug targets for distinctive drug design of Candida auris date: 2022-04-10 journal: bioRxiv DOI: 10.1101/2022.04.07.487516 sha: 57061dbf3a73e21c0ec6de6eaff76e85bcf0d0db doc_id: 297870 cord_uid: 4bw6egae Candida auris is a serious health concern of current world that possess serious global health threat and is emerging at a high rate. Available antifungal drugs are failing to combat this pathogen as they are growing resistance toward those drugs and some strains have already showed resistant to all three available antifungal drugs in the market. Finding alternative treatments is a must, therefore, to save lives from this foe. To make the way easier for developing new treatments, we have made some insilico analysis of this pathogen to identify suitable targets for designing drugs and also suggested some potential metabolites to test in vivo condition after some computational analysis. After the subtraction of duplicate, non-essential, human homologs, non-metabolic, human microbiome similar and druggable proteins we ended up with three proteins (XP_028890156.1, XP_028891672.1 and XP_028891858.1) from a total of 5441 C. auris proteins. Blocking those proteins will result in the destruction of the pathogen while the host will remain safe from unintentional blocking. Their subcellular locations and interaction with high number of proteins also indicate their suitability as drug target candidates. After analyzing in silico docking of 29 potential antifungal from plant origin with those three proteins we selected Caledonixanthone E, Viniferin, Glaucine, Jatrorrhizine as the most potent weapon to block those proteins as they showed higher binding affinity. Furthermore, they were predicted to be safe and also showed proper ADME properties (Figure 1). The once rare fungal pathogen C. auris drew the attention of world health community because of the increased mortality rate owing to severe invasive candidiasis and recurrent outbreaks since 2009 (Du et al., 2020; Jackson et al., 2019) . It was first detected from the ear canal discharge of a patient in Japan and hence introduced "auris" in its nomenclature. But it is thought to be present even long before its discovery as misdiagnosed human pathogen. The dermal sites, respiratory and urogenital tract and the blood circulatory system are the potential sites of fatal invasive C. auris infection in human body (Arensman et al., 2020; Hata, Humphries, & Lockhart, 2020) . It is a budding yeast (ovoid to elongate structure), which may form rudimentary pseudohyphae. However, these phenotypes are not distinct enough to assist the identification of C. auris. The whole genome sequencing, a VITEK 2 version 8.01 and mass spectrometry have shown promising results in the identifying and differentiating C. auris from other closely related Candida species and also in characterizing the dynamics of its transmission (Chow et al., 2020) . The unique traits of C. auris like; nosocomial transmission in clinical settings, misidentification, colonizing ability, multi-drug resistance (MDR), high virulence and mortality rates, long-term persistence on health care environmental surfaces (both on the human host and inanimate surfaces) separates it from other Candida species and justifies its potential to generate pandemic situation (Chow et al., 2020; Du et al., 2020; Forsberg et al., 2018; Welsh et al., 2017) . Besides the fungus shows high spread ability and can infect all kinds of patients irrespective of age and different immune-suppression diseases (Saris, Meis, & Voss, 2018) . Recent investigations over the COVID 19 conditions, indicates the higher probability of C. auris outbreaks in the ICUs of the overburdened hospitals. Hence extra caution is demanded regarding the identification and treatment of C. auris infections among COVID-19 patients (Anuradha Chowdhary & Sharma, 2020; Jackson et al., 2019) . The genome (haploid) of C. auris comprises of 7 chromosomes, having a genomic size ranging from 12.1 Mb to 12.7 Mb (with approximately 45% GC content). The analysis of the genome estimates about 6,500-8,500 protein-coding sequences and around 5,500 genes; that codes different virulent proteins in Candida species which develops virulent characters of the fungus like biofilm formation. The mortality rates at varied geographical locations portrayed significant differences. In Asia, the Far East, and the United States the mortality rate exceeds 50% whereas in Venezuela and and Colombia the rate is 72% and 35.2% respectively(W. G. Lee et al., 2011; Morales-López et al., 2017; Sarma et al., 2013; Vallabhaneni et al., 2017) . Till now three major classes of antifungal drugs i.e. azoles, polyenes and echinocandins are largely in practice regarding the treatment of invasive fungal infections or candidiasis. However, the multidrug resistance property of C. auris limits the options for the treatment of invasive infections caused by the pathogen. Different research groups and CDC (The Centers for Disease Control and Prevention) analyzed the resistance percentage of C. auris against different antifungals. For example; in a study 93% of the fungus among 54 internationally collected isolates showed resistance against fluconazole, voriconazole (>50%), amphotericin B (35%), echinocandins (7%) and two antifungal classes i.e. azoles and polyenes (41%). Besides 4% isolates demonstrated raised MICs (minimum inhibitory concentration) towards the 3 major antifungal classes and thus interpreting dangerous situation regarding the control and treatment of multidrug resistant fungal pathogen, C. auris (A. Chowdhary et al., 2013; Clancy & Nguyen, 2017; Lockhart et al., 2016) . Different side-effects and multi-drug resistance generated due to different antifungal treatments, points out the significance of novel drug discovery, evaluating the antifungal properties of diverse natural and synthetic/ semi-synthetic chemical compounds (Fuentefria, Pippi, Dalla Lana, Donato, & de Andrade, 2018) . The host proteins and genes, disease associated biomarkers, microRNSs and phenotypes, varied biological pathways and the connections among different biological networks and molecular functions are targeted by these compounds for generating treatment measures. Multiple computational and in silico methods are being used nowadays in assisting the drug discovery process and at the same time reducing the risk factors, time consumed, cost and raw-materials, unlike the traditional drug discovery approaches (Agamah et al., 2020; Zhong et al., 2018) . Besides this the computer aided drug design (CADD) methodologies limits the pharmacological practices on animal models, designs safe and novel drug candidates, and thus broadens the field for potential antifungal treatments (Brogi, Ramalho, Kuca, Medina-Franco, & Valko, 2020) . The multi-drug resistant condition and development of toxicity and side-effects in host, developed as a result of the current antifungal therapeutics application, stimulates the researches for the novel drug development. Therefore, in this study, subtractive proteomics approach was utilized to identify suitable drug targets of c. auris and reverse or inverse virtual screening or inverse docking technique was used that helped identifying the probable relationships among targeted proteins via chemical probing. The reverse docking approach screens different protein targets (ligands /compounds) in order to find the potent binding partners via statistical analysis, unlike the conventional forward docking where a variety of ligands docks to a particular target only (M. Lee & Kim, 2012) . The plant secondary metabolites having antimicrobial and antifungal activities are being tested for this purpose. More than 50% of the Food and Drug Administration (FDA) approved drugs are derived from the plant metabolites (Shital S Chavan, 2018) . This validates the wide range of therapeutic potential of the natural products to be used as the basis of the innovation of new drugs. The recent advancement and interference of bioinformatics and in silico screening have largely enhanced the proportion of detecting potent plant metabolites with therapeutic properties. For this purpose multiple computerized databases are established by the researchers containing detailed information about different plant metabolites. This information helps virtual screening and selection of the potential natural products as novel drugs against pathogen C. auris. Thus this study aims at screening the best drug target candidates and bioactive compounds from plant origin and utilizing them as the building blocks for the development of novel drugs against the deadly fungal pathogen C. auris. Several online server and bioinformatics tools were utilized to scrutinize the unique drug targets of Candida auris and potential drugs to block those unique targets. Representative Candida auris (assembly Cand_auris_B11221_V1) proteome was derived from the National Center for Biotechnology Information Genome database (https://www.ncbi.nlm.nih.gov/genome). To find paralog proteins, proteome file was submitted to the CD-HIT server (http://weizhonglab.ucsd.edu/cdhit_suite/cgi-bin/index. cgi?cmd=cd-hit) by setting "sequence identity cut-off" at 0.6 (60 percent identity) to eliminate repeated protein sequences (Huang, Niu, Gao, Fu, & Li, 2010) ; (Dutta et al., 2006) . Mini proteins (sequences less than 100 amino acids) play a crucial role in different regulatory activities and biological processes; hence the "length of sequence to skip" was set at 100 to exclude those (Gupta, Singh, Gupta, Kk, & Pk, 2010) . Amino acids having longer sequences, on the other hand, are frequently discovered to be engaged in important metabolic processes (Haag, Velk, & Wu, 2012) The Non-paralog protein sequences were subjected to BLAST against the DEG server database (www.essentialgene.org/) which is used to identify essential proteins (Luo, Lin, Gao, Zhang, & Zhang, 2014) . The cut-off value was set at e-value 1e-100 and per identity of 25%. To screen ortholog in the host, BLASTp was run against the human genome in the NCBI database for Non-paralog essential proteins. Except for Homo sapiens (taxid: 9606) in the "organisms" box, everything was kept the same, and the threshold e-value was set to 0.0001. Only queries with no significant hit were kept for further investigation with the hopes of developing a pathogen-specific remedy (Pourhajibagher & Bahador, 2016) . Both Candida auris and human metabolic pathways were obtained from the KEGG PATHWAY database (https://www.genome.jp/kegg/pathway.html) by entering their respective three-letter KEGG organism codes in the organism box: 'caur' for Candida auris and 'hsa' for H. sapiens (Ogata et al., 1999) .The bacterium's unique pathways were then screened by manual comparison. The KO number of critical host non homologous proteins was determined using BLASTp using the KEEG database's KAAS server (https://www.genome.jp/kaas-bin/kaas_main). The KO number was then used to search the KO server (KEGG ORTHOLOGY) (https://www.genome.jp/kegg/ko.html) for protein pathways, and it was determined which proteins are solely engaged in the pathogen's unique metabolic pathways (Damte et al., 2013) . Proteins determined to be involved in common pathways were excluded from the study. The human body has a huge number of beneficial microbes that defend the body against diseases and dangerous foreign particles. The goal of the non-homology study of the human microbiome was to see if the unique metabolic pathway sequences of Candida auris matches to the protein sequences of those beneficial bacteria found in the human body. Therefore, unique metabolic pathway protein sequences were subjected to BLAST against Bioproject-43021 in the NCBI BLAST software (Turnbaugh et al., 2007) and 0.005 was chosen as the cut-off score. Sequences with a similarity of less than 45 percent are maintained for further steps. Chosen proteins from the previous steps were searched in the DrugBank database (https://www.drugbank.ca/structures/search/bonds/sequence) to reveal the uniqueness of the selected proteins as drug targets (Wishart et al., 2018) . The existence of targets indicates their draggability, but the absence of targets indicates the uniqueness of the proteins classed as "new targets" (Knox et al., 2011) . During the action, all of the parameters were left at their default values. Membrane proteins can be exploited as both drug targets and vaccination candidates in cases where cytoplasmic proteins have the ability to function as drug targets (Mahmud, Khan, & Iqbal, 2019) . CELLO v.2.5 (http://cello.life.nctu.edu.tw/) server was utilized to forecast the subcellular localization of the unique drug targets (Yu, Chen, Lu, & Hwang, 2006) . The protein-protein interaction network of the unique drug targets was analyzed using the STRING 11.5 (https://string-db.org/) server (Szklarczyk et al., 2019) . To eliminate misleading negative and positive results, the Protein network includes high confidence interactors with scores ≥ 0.700. The number of interacting proteins (nodes) and interactions (edges) indicate when the query protein will be removed what effects it will put on fungal metabolic system (Kushwaha & Shakya, 2010) . Protein Data Bank did not provide structures for the proteins chosen (RCSB PDB). So the I-TASSER service was used to predict molecular models of unique proteins (Roy, Kucukural, & Zhang, 2010) , which were then refined using the GalaxyWEB server (http://galaxy.seoklab.org/cgi-bin/submit.cgi?type=REFINE) (Ko, Park, Heo, & Seok, 2012) . Saves server v6.0 (https://saves.mbi.ucla.edu/) was used to analyze the Errat quality score and ramachandan plot of the models to identify the best model and RaptorX Binding site (http://raptorx.uchicago.edu/BindingSite/) prediction server was used to estimate ligand binding sites of the proteins (Colovos & Yeates, 1993) ; (Laskowski, Rullmannn, MacArthur, Kaptein, & Thornton, 1996) . . The interaction between tiny ligands and macromolecules may be predicted using molecular docking in drug development (Kitchen, Decornez, Furr, & Bajorath, 2004 analysis. The docked complexes were then refined using the FireDock refinement program (Mashiach, Schneidman-Duhovny, Andrusier, Nussinov, & Wolfson, 2008) . The PyMOL v2.0 program was then used to visualize and analyze binding sites of the metabolites (Wang et al., 2015) . The features of adsorption, distribution, metabolism, and excretion (ADME) are primarily linked to the kinetics of drug exposure to tissue. Analyzing ADME during the discovery phase will lower the chance of pharmacokinetics-related clinical failure (Hay, Thomas, Craighead, Economides, & Rosenthal, 2014) . The ADME characteristics of the top four metabolites were evaluated using the SwissADME server (Daina, Michielin, & Zoete, 2017) . The medications were uploaded in SDF format to the server, transformed to SMILES, and then run to obtain predictions. The BOILED-Egg model was then used to determine the blood-brain barrier (BBB) of the substances examined (Daina & Zoete, 2016) . To forecast the comparative harmful effects of leading medications, researchers used pkCSM, an online program (Pires, Blundell, & Ascher, 2015) that requires SMILES for the top four metabolites, which were obtained from PubChem database. NCBI has a total of 112 genomes for Candida auris, from which we compiled the Candida auris representative proteome (assembly Cand_auris_B11221_V1), which contained 5441 proteins in total. At 60% identity, the CD-HIT server discovered 5217 clusters therefore, after removing the paralogous sequences, 5217 non-duplicate large proteins remained (Supplementary file 1). Proteins for which significant hit was found under the selected conditions were considered essential proteins for Candida auris. 795 out of 5217 non-paralog proteins were found to be essential for the organism which were selected for further analysis (Supplementary file 2) . Drugs and therapeutic compounds should be designed in such a way that they do not unintentionally block host proteins. Therefore, human non-homology step is performed to reduce unfavorable drug binding to the active sites of the host homologous proteins (Sarkar, Maganti, Ghoshal, & Dutta, 2012) . As a result, orthologs were taken off the list. In this step, 66 non-paralog essential proteins have no hits with human proteins which were regarded as non-orthologous to the host (Supplementary file 3) . The KEGG server had 122 C. auris metabolic pathways and 345 human metabolic pathways, with 18 metabolic pathways (Supplementary file 4) being specific for C. auris. Proteins involved in these unique pathways can be used as therapeutic targets. Following a BLAST at the KAAS server, 52 out of 66 non-homolog essential proteins were discovered to have both KO orthology and metabolic pathway involvement. These 52 proteins are important for the smooth operation of metabolism, and 19 of them were found to be solely associated with C. auris unique pathways (Table-1) . Microbiome analysis revealed that 6 pathogen proteins (Supplementary file 5) shared less than 45 percent similarity with human microflora, three of which shared similarity with approved, investigational, and experimental drug targets of Drugbank Database (Supplementary file 6). We ended up taking these three proteins off the list because we wanted to find novel targets. Three proteins did not show any resemblance as a novel targets. As a result, they were nominated (Table 2) as the unique drug targets of the pathogen. All of the unique pathway proteins are present either in the cytoplasm or in the inner membrane. XP 028890156.1 and XP 028891672.1 are plasma membrane proteins, while XP 028891858.1 is found in the cytoplasm (Table-2) . We kept all of them for further analysis because both cytoplasmic and membrane proteins can be used as potential drug targets. Proteins that interact with multiple proteins are thought to be metabolically active and they are suitable as drug target (Cui, Zhang, Wang, & He, 2009; Kushwaha & Shakya, 2010) . STRING revealed that XP_028890156.1, XP_028891672.1 and XP_028891858.1 each exhibit interactions with 10 proteins (Figure 2 ). XP_028891858.1 with other proteins where red color balls represent the target proteins. Ramachandan plot, the best one was then refined via GalaxyWEB server. This server provided 5 refined models for each protein. Finally, we took the best model by analyzing their ERRAT quality score and Ramachandan plot analysis. ERRAT value, Ramachandan plot result and the binding sites residues predicted by RaptorX Binding site server are provided in Table 3 , Figure 3 , Figure 4 and Figure 5 . To characterize drug profiles of top antibacterial drugs, ADME properties of top drugs were evaluated ( Table 5 ). All of the metabolites had a high rate of absorption in the gastrointestinal tract. In the case of Caledonixanthone E and Viniferin, no BBB was found in the top drugs. Furthermore, all of the drugs showed moderate level of water solubility except for Viniferin and no metabolite had indicated that it may induce pain (Table 5) . The toxicity profiles of top metabolites are enlisted in Table 6 . The LD50 values for the top drugs ranged from 2.131 to 3.219 mol/kg and they displayed negative skin sensitization result and oral rat acute toxicity. Minnow Toxicity values of all drugs were more than -0.3 log mM, proving them non-toxic. Again, negative hepatotoxicity results, except for Jatrorrhizine, of all drugs indicate that the normal function of the liver will not be disrupted through these top drugs. Essential proteins which are deemed as most appropriate antimicrobial therapeutic targets since the majority of the drugs have a tendency to dock along with essential gene products. The whole emblematic proteome of Candida auris, comprised of 5441 proteins, in addition to, the paralog proteins (60% matching) were too excluded as the motifs, domains, and active site etc. might be comparable for paralogs. A prospective drug target have to be an essential protein as it has a vital attribute for the existence of the pathogen (Sarkar et al., 2012) and our target pathogen yeast includes 795 such essential proteins specifying that drugs that will act out in opposition to those proteins will lead the pathogen to death. Subtraction of hosthomolog proteins is regarded as the notable step of in-silico drug target determination (Anishetty, Pulimi, & Pennathur, 2005; Sarkar et al., 2012) . After exploration, 66 such nonorthologs were obtained against which potential drugs can be devised that may not trigger cross reaction (Sarkar et al., 2012) . Amid those non-orthologs, 52 proteins were anticipated to be linked with the metabolism of the yeast but amongst them, only 19 were detected related only with the unique C. auris metabolic pathways (Table 1) . Assessment of host micro biome non-similarity makes sure that the defense of those useful microbes from unintentional blocking. Consequently, 13 distinctive metabolic pathway proteins demonstrating more correspondence with the beneficial microbes were removed. The Drug-Bank database was inspected for evading mutational alterations as well as to prevent the progress of sturdy bacteria by the broad-spectrum therapeutic drugs. Among the six microbiome non-similar proteins, three were druggable and may possibly be disabled by already permitted and accessible drugs. Hence, the remaining three proteins were retained as the novel drug targets (Table 2) . Although proteins are uncovered to be confined in five diverse site of the calls, membrane and cytoplasmic one is regarded as the elite for drug targets (Michael, Dominey-Howes, & Labbate, 2014) . Every distinctive protein are situated either in the cytoplasm or in the membrane showing their capability to act as a drug target (Table 2) . Protein-protein interactions (PPIs) for all the distinctive proteins were discovered ( Figure 2 ) as it allocates us to comprehend the function of a protein in a precise pathway and changes or mutations in the protein-protein network interrupt the regular flow of activities of the cell (genengnews, 2005, March 5). All the three proteins that were found to intermingle with 10 proteins demonstrating the incidental function interruption of those 10 proteins by the obstruction of the drug targets. Plant metabolites play a significant part by being a principal molecule in unearthing appropriate drug contenders (Joseph, Bhaskaran, Kaliraj, Muthuswamy, & Suresh, 2017) . And so, we considered some inhibitory metabolites derived from plant sources for C. auris on the basis of their affinity of joining to the certain unique drug targets. The outcome of docking has shown that four drug molecules i.e. Caledonixanthone E, Viniferin, Glaucine and Jatrorrhizine have high affinity for all of the three macromolecules with the least global binding energy. On the other hand, Viniferin exhibited maximum binding affinity with each of the three drug targets (Table 4 ). The structural configurations of the docked protein complex were scrutinized to interpret the exterior drug hotspot of the targeted distinctive proteins. The binding arrangement of ligand and interrelating residues with their corresponding locations was inspected (Table 4 ) ( Figure 6, 7 and 8 ). This inspection leads to the finding of the amino acids from 02-21 and 276-322 positions that were fundamental to the binding interactions of >XP_028890156.1. However, residues 9-26 were discovered as essential for the binding of >XP_028891672.1, the binding hotspot of ligands for protein >XP_028891858.1 that lies between 53 and 80 positions. Inadequate ADME data which is frequently linked with malfunction of clinical trials in the course of several drug development projects (Shin, Kang, & No, 2016) . Thus, ADME data analysis is crucial for those types of projects which can be accomplished by in vitro or in vivo or in-silico approaches. The main four drug contenders demonstrated no adverse effects in in-silico ADME study which may perhaps lessen the drug related properties ( Table 5) . Each of the potential drugs displayed water solubility and absorption in GI. The prediction of toxicity revealed that all of the four drug contenders are non-carcinogenic, non-mutagenic and impervious to skin and nonhepatotoxic. On the whole, the toxicity analysis of the drugs made known that the predicted novel drugs are harmless to conduct operations and can be used as therapeutic medications to treat C. auris Finally selected three proteins and four metabolites to block those three proteins may be greater foresteps of drug discovery against the emerging pathogen, C. auris. This study will speed up the process to get remedies with less trials and error repeats of assays. However, in vivo and clinical trials are highly recommended to validate the final outcome of the study. Md. Nazmul Islam Bappy: Idea generation, conceptualization, experiment design, data handling, data analysis, manuscript writing and draft preparation. Kazi Md. Ali Zinnah: Supervision of experimental design, project administration, manuscript writing, reviewing and approval of final manuscript Tanjin Barketullah Robin, Anindita Ash Prome, Fayeza Sadia Laskar, Anindita Roy, Hafsa Akter: Data handling, data analysis and manuscript writing Dataset of this study is available from the corresponding author on reasonable request The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors Computational/in silico methods in drug target and lead prediction Potential drug targets in Mycobacterium tuberculosis through metabolic pathway analysis Clinical Outcomes of Patients Treated for Candida auris Infections in a Multisite Health System Editorial: In silico Methods for Drug Design and Discovery. Frontiers in Chemistry Tracing the Evolutionary History and Global Expansion of Candida auris Using Population Genomic Analyses The lurking scourge of multidrug resistant Candida auris in times of COVID-19 pandemic New clonal strain of Candida auris Emergence of Candida auris: An International Call to Arms Verification of protein structures: patterns of nonbonded atomic interactions Uncovering new signaling proteins and potential drug targets through the interactome analysis of Mycobacterium tuberculosis SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules A BOILED-Egg To Predict Gastrointestinal Absorption and Brain Penetration of Small Molecules Putative drug and vaccine target protein identification using comparative genomic analysis of KEGG annotated metabolic pathways of Mycoplasma hyopneumoniae Candida auris: Epidemiology, biology, antifungal resistance, and virulence In silico identification of potential therapeutic targets in the human pathogen Helicobacter pylori Candida auris: The recent emergence of a multidrug-resistant fungal pathogen Antifungals discovery: an insight into new strategies to combat antifungal resistance Protein-Protein Interactions in Drug Discovery Definition of Potential Targets in Mycoplasma Pneumoniae Through Subtractive Genome Analysis Potential Antibacterial Targets in Bacterial Central Metabolism Candida auris: An Emerging Yeast Pathogen Posing Distinct Challenges for Laboratory Diagnostics, Treatment, and Infection Prevention CD-HIT Suite: a web server for clustering and comparing biological sequences On the Origins of a Species: What Might Explain the Rise of Candida auris? Molecular Docking of Phytoligands to the viral protein receptor P. monodon Rab7 PubChem Substance and Compound databases Docking and scoring in virtual screening for drug discovery: methods and applications DrugBank 3.0: a comprehensive resource for 'omics' research on drugs GalaxyWEB server for protein structure prediction and refinement Protein interaction network analysis--approach for potential drug target identification in Mycobacterium tuberculosis AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR Large-scale reverse docking profiles and their applications First three reported cases of nosocomial fungemia caused by Candida auris Simultaneous Emergence of Multidrug-Resistant Candida auris on 3 Continents Confirmed by Whole-Genome Sequencing and Epidemiological Analyses DEG 10, an update of the database of essential genes that includes both protein-coding genes and noncoding genomic elements Identification of novel drug targets for humans and potential vaccine targets for cattle by subtractive genomic analysis of Brucella abortus strain 2308 FireDock: a web server for fast interaction refinement in molecular docking Molecular docking: a powerful approach for structure-based drug discovery The antimicrobial resistance crisis: causes, consequences, and management. Frontiers in public health Invasive Infections with Multidrug-Resistant Yeast Candida auris, Colombia. Emerging infectious diseases Open Babel: An open chemical toolbox KEGG: Kyoto Encyclopedia of Genes and Genomes pkCSM: Predicting Small-Molecule Pharmacokinetic and Toxicity Properties Using Graph-Based Signatures Designing and in Silico Analysis of PorB Protein from Chlamydia Trachomatis for Developing a Vaccine Candidate I-TASSER: a unified platform for automated protein structure and function prediction Candida auris In silico quest for putative drug targets in Helicobacter pylori HPAG1: molecular modeling of candidate enzymes from lipopolysaccharide biosynthesis pathway Candidemia caused by amphotericin B and fluconazole resistant Candida auris PatchDock and SymmDock: servers for rigid and symmetric docking Predicting ADME Properties of Chemicals Antibacterial and Antifungal Drugs from Natural Source: A Review of Clinical Development STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets The Human Microbiome Project Investigation of the First Seven Reported Cases of Candida auris, a Globally Emerging Invasive, Multidrug-Resistant Fungus-United States Interaction of α -cyperone with human serum albumin: Determination of the binding site by using Discovery Studio and via spectroscopic methods Survival, Persistence, and Isolation of the Emerging Multidrug-Resistant Pathogenic Yeast Candida auris on a Plastic Health Care Surface DrugBank 5.0: a major update to the DrugBank database Prediction of protein subcellular localization Artificial intelligence in drug design