Machine Learning-based Tools Research Articles

Antihypertensive and Antidiabetic Drug Candidates from Milkfish (Chanos chanos)-Identification and Characterization through an Integrated Bioinformatic Approach.

Integrated bioinformatics tools have created more efficient and robust methods to overcome in vitro challenges and have been widely utilized for the investigation of food proteins and the generation of peptide sequences. This study aimed to analyze the physicochemical properties and bioactivities of novel peptides derived from hydrolyzed milkfish (Chanos chanos) protein sequences and to discover their potential angiotensin-converting enzyme (ACE)- and dipeptidyl peptidase-4 (DPPIV)-inhibitory activities using machine learning-based tools, including BIOPEP-UWM, PeptideRanker, and the molecular docking software HADDOCK 2.4. Nine and three peptides were predicted to have ACE- and DPPIV-inhibitory activities, respectively. The DPPIV-inhibitory peptides were predicted to inhibit the compound with no known specific mode. Meanwhile, two tetrapeptides (MVWH and PPPS) were predicted to possess a competitive mode of ACE inhibition by directly binding to the tetra-coordinated Zn ion. Among all nine discovered ACE-inhibitory peptides, only the PPPS peptide satisfied the drug-likeness analysis requirements with no violations of the Lipinski rule of five and should be further investigated in vitro.

Accuracy of renovo predictions on variants reclassified over time

BackgroundInterpreting the clinical consequences of genetic variants is the central problem in modern clinical genomics, for both hereditary diseases and oncology. However, clinical validation lags behind the pace of discovery, leading to distressing uncertainty for patients, physicians and researchers. This “interpretation gap” changes over time as evidence accumulates, and variants initially deemed of uncertain (VUS) significance may be subsequently reclassified in pathogenic/benign. We previously developed RENOVO, a random forest-based tool able to predict variant pathogenicity based on publicly available information from GnomAD and dbNFSP, and tested on variants that have changed their classification status over time. Here, we comprehensively evaluated the accuracy of RENOVO predictions on variants that have been reclassified over the last four years.Methodswe retrieved 16 retrospective instances of the ClinVar database, every 3 months since March 2020 to March 2024, and analyzed time trends of variant classifications. We identified variants that changed their status over time and compared RENOVO predictions generated in 2020 with the actual reclassifications.ResultsVUS have become the most represented class in ClinVar (44.97% vs. 9.75% (likely) pathogenic and 40,33% (likely) benign). The rate of VUS reclassification is linear and slow compared to the rate of VUS reporting, exponential and currently ~ 30x faster, creating a growing divide between what can be sequenced vs. what can be interpreted. Out of 10,196 VUS variants in January 2020 that have undergone a clinically meaningful reclassification to march 2024, RENOVO correctly classified 82.6% in 2020. In addition, RENOVO correctly identified the majority of the few variants that switched clinically meaningful classes (e.g., from benign to pathogenic and vice versa). We highlight variant classes and clinically relevant genes for which RENOVO provides particularly accurate estimates. In particularly, genes characterized by large prevalence of high- or low-impact variants (e.g., POLE, NOTCH1, FANCM etc.). Suboptimal RENOVO predictions mostly concern genes validated through dedicated consortia (e.g., BRCA1/2), in which RENOVO would anyway have a limited impact.ConclusionsTime trend analysis demonstrates that the current model of variant interpretation cannot keep up with variant discovery. Machine learning-based tools like RENOVO confirm high accuracy that can aid in clinical practice and research.

Monitoring nucleolar-nucleoplasmic protein shuttling in living cells by high-content microscopy and automated image analysis.

The nucleolus has corefunctions in ribosome biosynthesis, but also acts as a regulatory hub in a plethora of non-canonical processes, including cellular stress. Upon DNA damage, several DNA repair factors shuttle between the nucleolus and the nucleoplasm. Yet, the molecular mechanisms underlying such spatio-temporal protein dynamics remain to be deciphered. Here, we present a novel imaging platform to investigate nucleolar-nucleoplasmic protein shuttling in living cells. For image acquisition, we used a commercially available automated fluorescence microscope and for image analysis, we developed a KNIME workflow with implementation of machine learning-based tools. We validated the method with different nucleolar proteins, i.e., PARP1, TARG1 and APE1, by monitoring their shuttling dynamics upon oxidative stress. As a paradigm, we analyzed PARP1 shuttling upon H2O2 treatment in combination with a range of pharmacological inhibitors in a novel reporter cell line. These experiments revealed that inhibition of SIRT7 results in a loss of nucleolar PARP1 localization. Finally, we unraveled specific differences in PARP1 shuttling dynamics after co-treatment with H2O2 and different clinical PARP inhibitors. Collectively, this work delineates a highly sensitive and versatile bioimaging platform to investigate swift nucleolar-nucleoplasmic protein shuttling in living cells, which can be employed for pharmacological screening and in-depth mechanistic analyses.

Transformations establishing equivalence across neural networks: When have two networks learned the same task?

Transformations are a key tool in the qualitative study of dynamical systems: transformations to a normal form, for example, underpin the study of instabilities and bifurcations. In this work, we test, and when possible establish, an equivalence between two different artificial neural networks by attempting to construct a data-driven transformation between them, using diffusion maps with a Mahalanobis-like metric. If the construction succeeds, the two networks can be thought of as belonging to the same equivalence class. We first discuss transformation functions between only the outputs of the two networks; we then also consider transformations that take into account outputs (activations) of a number of internal neurons from each network. Whitney's theorem dictates the number of (generic) measurements from one of the networks required to reconstruct each and every feature of the second network. The construction of the transformation function relies on a consistent, intrinsic representation of the network input space. We illustrate our algorithm by matching neural network pairs trained to learn (a) observations of scalar functions, (b) observations of two-dimensional vector fields, and (c) representations of images of a moving three-dimensional object (a rotating horse). We also demonstrate reconstruction of a network's input (and output) from minimal partial observations of intermediate neuron activations. The construction of equivalences across different network instantiations clearly relates to transfer learning and will also be valuable in establishing equivalence between different machine learning-based tools.

Predicting acute kidney injury onset with a random forest algorithm using electronic medical records of COVID-19 patients: the CRACoV-AKI model.

Acute kidney injury (AKI) is a serious and common complication of SARS‑CoV‑2 infection. Most risk assessment tools for AKI have been developed in the intensive care unit or in elderly populations. As the COVID‑19 pandemic is transitioning into an endemic phase, there is an unmet need for prognostic scores tailored to the population of patients hospitalized for this disease. We aimed to develop a robust predictive model for the occurrence of AKI in hospitalized patients with COVID‑19. Electronic medical records of all adult inpatients admitted between March 2020 and January 2022 were extracted from the database of a large, tertiary care center with a reference status in Lesser Poland. We screened 5806 patients with SARS‑CoV‑2 infection confirmed with a polymerase chain reaction test. After excluding individuals with lacking data on serum creatinine levels and those with a mild disease course (<7 days of inpatient care), a total of 4630 records were considered. Data were randomly split into training (n = 3462) and test (n = 1168) sets. A random forest model was tuned with feature engineering based on expert advice and metrics evaluated in nested cross‑validation to reduce bias. Nested cross‑validation yielded an area under the curve ranging between 0.793 and 0.807, and an average performance of 0.798. Model explanation techniques from a global perspective suggested that a need for respiratory support, chronic kidney disease, and procalcitonin concentration were among the most important variables in permutation tests. The CRACoV‑AKI model enables AKI risk stratification among hospitalized patients with COVID‑19. Machine learning-based tools may thus offer additional decision‑making support for specialist providers.

Interpretation of the Visual Field in Neuro-ophthalmic Disorders.

In this review, we will describe current methods for visual field testing in neuro-ophthalmic clinical practice and research, develop terminology that accurately describes patterns of field deficits, and discuss recent advances such as augmented or virtual reality-based perimetry and the use of artificial intelligence in visual field interpretation. New testing strategies that reduce testing times, improve patient comfort, and increase sensitivity for detecting small central or paracentral scotomas have been developed for static automated perimetry. Various forms of machine learning-based tools such as archetypal analysis are being tested to quantitatively depict and monitor visual field abnormalities in optic neuropathies. Studies show that the combined use of optical coherence tomography and standard automated perimetry to determine the structure-function relationship improves clinical care in neuro-ophthalmic disorders. Visual field assessment must be performed in all patients with neuro-ophthalmic disorders affecting the afferent visual pathway. Quantitative visual field analysis using standard automated perimetry is critical in initial diagnosis, monitoring disease progression, and guidance of therapeutic plans. Visual field defects can adversely impact activities of daily living such as reading, navigation, and driving and thus impact quality of life. Visual field testing can direct appropriate occupational low vision rehabilitation in affected individuals.

Sepsis Prediction at Emergency Department Triage Using Natural Language Processing: Retrospective Cohort Study.

Despite its high lethality, sepsis can be difficult to detect on initial presentation to the emergency department (ED). Machine learning-based tools may provide avenues for earlier detection and lifesaving intervention. The study aimed to predict sepsis at the time of ED triage using natural language processing of nursing triage notes and available clinical data. We constructed a retrospective cohort of all 1,234,434 consecutive ED encounters in 2015-2021 from 4 separate clinically heterogeneous academically affiliated EDs. After exclusion criteria were applied, the final cohort included 1,059,386 adult ED encounters. The primary outcome criteria for sepsis were presumed severe infection and acute organ dysfunction. After vectorization and dimensional reduction of triage notes and clinical data available at triage, a decision tree-based ensemble (time-of-triage) model was trained to predict sepsis using the training subset (n=950,921). A separate (comprehensive) model was trained using these data and laboratory data, as it became available at 1-hour intervals, after triage. Model performances were evaluated using the test (n=108,465) subset. Sepsis occurred in 35,318 encounters (incidence 3.45%). For sepsis prediction at the time of patient triage, using the primary definition, the area under the receiver operating characteristic curve (AUC) and macro F1-score for sepsis were 0.94 and 0.61, respectively. Sensitivity, specificity, and false positive rate were 0.87, 0.85, and 0.15, respectively. The time-of-triage model accurately predicted sepsis in 76% (1635/2150) of sepsis cases where sepsis screening was not initiated at triage and 97.5% (1630/1671) of cases where sepsis screening was initiated at triage. Positive and negative predictive values were 0.18 and 0.99, respectively. For sepsis prediction using laboratory data available each hour after ED arrival, the AUC peaked to 0.97 at 12 hours. Similar results were obtained when stratifying by hospital and when Centers for Disease Control and Prevention hospital toolkit for adult sepsis surveillance criteria were used to define sepsis. Among septic cases, sepsis was predicted in 36.1% (1375/3814), 49.9% (1902/3814), and 68.3% (2604/3814) of encounters, respectively, at 3, 2, and 1 hours prior to the first intravenous antibiotic order or where antibiotics where not ordered within the first 12 hours. Sepsis can accurately be predicted at ED presentation using nursing triage notes and clinical information available at the time of triage. This indicates that machine learning can facilitate timely and reliable alerting for intervention. Free-text data can improve the performance of predictive modeling at the time of triage and throughout the ED course.

Cybersecurity in Supply Chain Systems: The Farm-to-Fork Use Case

Modern supply chains comprise an increasing number of actors which deploy different information technology systems that capture information of a diverse nature and diverse sources (from sensors to order information). While the benefits of the automatic exchange of information between these systems have been recognized and have led to their interconnection, protecting the whole supply chain from potential attacks is a challenging issue given the attack proliferation reported in the literature. In this paper, we present the FISHY platform, which anticipates protecting the whole supply chain from potential attacks by (a) adopting novel technologies and approaches including machine learning-based tools to detect security threats and recommend mitigation policies and (b) employing blockchain-based tools to provide evidence of the captured events and suggested policies. This platform is also easily expandable to protect against additional attacks in the future. We experiment with this platform in the farm-to-fork supply chain to prove its operation and capabilities. The results show that the FISHY platform can effectively be used to protect the supply chain and offers high flexibility to its users.

Emergent Applications of Machine Learning for Diagnosing and Managing Appendicitis: A State-of-the-Art Review.

Background: Appendicitis is an inflammatory condition that requires timely and effective intervention. Despite being one of the most common surgically treated diseases, the condition is difficult to diagnose because of atypical presentations. Ultrasound and computed tomography (CT) imaging improve the sensitivity and specificity of diagnoses, yet these tools bear the drawbacks of high operator dependency and radiation exposure, respectively. However, new artificial intelligence tools (such as machine learning) may be able to address these shortcomings. Methods: We conducted a state-of-the-art review to delineate the various use cases of emerging machine learning algorithms for diagnosing and managing appendicitis in recent literature. The query ("Appendectomy" OR "Appendicitis") AND ("Machine Learning" OR "Artificial Intelligence") was searched across three databases for publications ranging from 2012 to 2022. Upon filtering for duplicates and based on our predefined inclusion criteria, 39 relevant studies were identified. Results: The algorithms used in these studies performed with an average accuracy of 86% (18/39), a sensitivity of 81% (16/39), a specificity of 75% (16/39), and area under the receiver operating characteristic curves (AUROCs) of 0.82 (15/39) where reported. Based on accuracy alone, the optimal model was logistic regression in 18% of studies, an artificial neural network in 15%, a random forest in 13%, and a support vector machine in 10%. Conclusions: The identified studies suggest that machine learning may provide a novel solution for diagnosing appendicitis and preparing for patient-specific post-operative complications. However, further studies are warranted to assess the feasibility and advisability of implementing machine learning-based tools in clinical practice.

Application of artificial intelligence approaches to predict the metabolism of xenobiotic molecules by human gut microbiome.

A highly complex, diverse, and dense community of more than 1,000 different gut bacterial species constitutes the human gut microbiome that harbours vast metabolic capabilities encoded by more than 300,000 bacterial enzymes to metabolise complex polysaccharides, orally administered drugs/xenobiotics, nutraceuticals, or prebiotics. One of the implications of gut microbiome mediated biotransformation is the metabolism of xenobiotics such as medicinal drugs, which lead to alteration in their pharmacological properties, loss of drug efficacy, bioavailability, may generate toxic byproducts and sometimes also help in conversion of a prodrug into its active metabolite. Given the diversity of gut microbiome and the complex interplay of the metabolic enzymes and their diverse substrates, the traditional experimental methods have limited ability to identify the gut bacterial species involved in such biotransformation, and to study the bacterial species-metabolite interactions in gut. In this scenario, computational approaches such as machine learning-based tools presents unprecedented opportunities and ability to predict the gut bacteria and enzymes that can potentially metabolise a candidate drug. Here, we have reviewed the need to identify the gut microbiome-based metabolism of xenobiotics and have provided comprehensive information on the available methods, tools, and databases to address it along with their scope and limitations.

Combination of expert guidelines-based and machine learning-based approaches leads to superior accuracy of automated prediction of clinical effect of copy number variations

Clinical interpretation of copy number variants (CNVs) is a complex process that requires skilled clinical professionals. General recommendations have been recently released to guide the CNV interpretation based on predefined criteria to uniform the decision process. Several semiautomatic computational methods have been proposed to recommend appropriate choices, relieving clinicians of tedious searching in vast genomic databases. We have developed and evaluated such a tool called MarCNV and tested it on CNV records collected from the ClinVar database. Alternatively, the emerging machine learning-based tools, such as the recently published ISV (Interpretation of Structural Variants), showed promising ways of even fully automated predictions using broader characterization of affected genomic elements. Such tools utilize features additional to ACMG criteria, thus providing supporting evidence and the potential to improve CNV classification. Since both approaches contribute to evaluation of CNVs clinical impact, we propose a combined solution in the form of a decision support tool based on automated ACMG guidelines (MarCNV) supplemented by a machine learning-based pathogenicity prediction (ISV) for the classification of CNVs. We provide evidence that such a combined approach is able to reduce the number of uncertain classifications and reveal potentially incorrect classifications using automated guidelines. CNV interpretation using MarCNV, ISV, and combined approach is available for non-commercial use at https://predict.genovisio.com/.

On the importance of interpretable machine learning predictions to inform clinical decision making in oncology.

Machine learning-based tools are capable of guiding individualized clinical management and decision-making by providing predictions of a patient's future health state. Through their ability to model complex nonlinear relationships, ML algorithms can often outperform traditional statistical prediction approaches, but the use of nonlinear functions can mean that ML techniques may also be less interpretable than traditional statistical methodologies. While there are benefits of intrinsic interpretability, many model-agnostic approaches now exist and can provide insight into the way in which ML systems make decisions. In this paper, we describe how different algorithms can be interpreted and introduce some techniques for interpreting complex nonlinear algorithms.

CRISPR genome editing using computational approaches: A survey.

Clustered regularly interspaced short palindromic repeats (CRISPR)-based gene editing has been widely used in various cell types and organisms. To make genome editing with Clustered regularly interspaced short palindromic repeats far more precise and practical, we must concentrate on the design of optimal gRNA and the selection of appropriate Cas enzymes. Numerous computational tools have been created in recent years to help researchers design the best gRNA for Clustered regularly interspaced short palindromic repeats researches. There are two approaches for designing an appropriate gRNA sequence (which targets our desired sites with high precision): experimental and predicting-based approaches. It is essential to reduce off-target sites when designing an optimal gRNA. Here we review both traditional and machine learning-based approaches for designing an appropriate gRNA sequence and predicting off-target sites. In this review, we summarize the key characteristics of all available tools (as far as possible) and compare them together. Machine learning-based tools and web servers are believed to become the most effective and reliable methods for predicting on-target and off-target activities of Clustered regularly interspaced short palindromic repeats in the future. However, these predictions are not so precise now and the performance of these algorithms -especially deep learning one's-depends on the amount of data used during training phase. So, as more features are discovered and incorporated into these models, predictions become more in line with experimental observations. We must concentrate on the creation of ideal gRNA and the choice of suitable Cas enzymes in order to make genome editing with Clustered regularly interspaced short palindromic repeats far more accurate and feasible.

Improvement of Mucosal Lesion Diagnosis with Machine Learning Based on Medical and Semiological Data: An Observational Study.

Despite artificial intelligence used in skin dermatology diagnosis is booming, application in oral pathology remains to be developed. Early diagnosis and therefore early management, remain key points in the successful management of oral mucosa cancers. The objective was to develop and evaluate a machine learning algorithm that allows the prediction of oral mucosa lesions diagnosis. This cohort study included patients followed between January 2015 and December 2020 in the oral mucosal pathology consultation of the Toulouse University Hospital. Photographs and demographic and medical data were collected from each patient to constitute clinical cases. A machine learning model was then developed and optimized and compared to 5 models classically used in the field. A total of 299 patients representing 1242 records of oral mucosa lesions were used to train and evaluate machine learning models. Our model reached a mean accuracy of 0.84 for diagnostic prediction. The specificity and sensitivity range from 0.89 to 1.00 and 0.72 to 0.92, respectively. The other models were proven to be less efficient in performing this task. These results suggest the utility of machine learning-based tools in diagnosing oral mucosal lesions with high accuracy. Moreover, the results of this study confirm that the consideration of clinical data and medical history, in addition to the lesion itself, appears to play an important role.

Sensitivity estimation for dark matter subhalos in synthetic Gaia DR2 using deep learning

The abundance of dark matter subhalos orbiting a host galaxy is a generic prediction of the cosmological framework, and is a promising way to constrain the nature of dark matter. In this paper, we investigate the use of machine learning-based tools to quantify the magnitude of phase-space perturbations caused by the passage of dark matter subhalos. A simple binary classifier and an anomaly detection model are proposed to estimate if stars or star particles close to dark matter subhalos are statistically detectable in simulations. The simulated datasets are three Milky Way-like galaxies and nine synthetic Gaia DR2 surveys derived from these. Firstly, we find that the anomaly detection algorithm, trained on a simulated galaxy with full 6D kinematic observables and applied on another galaxy, is nontrivially sensitive to the dark matter subhalo population. On the other hand, the classification-based approach is not sufficiently sensitive due to the extremely low statistics of signal stars for supervised training. Finally, the sensitivity of both algorithms in the Gaia-like surveys is negligible. The enormous size of the Gaia dataset motivates the further development of scalable and accurate data analysis methods that could be used to select potential regions of interest for dark matter searches to ultimately constrain the Milky Way’s subhalo mass function, as well as simulations where to study the sensitivity of such methods under different signal hypotheses.

Convolutional Neural Network-Based Compound Fingerprint Prediction for Metabolite Annotation.

Metabolite annotation has been a challenging issue especially in untargeted metabolomics studies by liquid chromatography coupled with mass spectrometry (LC-MS). This is in part due to the limitations of publicly available spectral libraries, which consist of tandem mass spectrometry (MS/MS) data acquired from just a fraction of known metabolites. Machine learning provides the opportunity to predict molecular fingerprints based on MS/MS data. The predicted molecular fingerprints can then be used to help rank putative metabolite IDs obtained by using either the precursor mass or the formula of the unknown metabolite. This method is particularly useful to help annotate metabolites whose corresponding MS/MS spectra are missing or cannot be matched with those in accessible spectral libraries. We investigated a convolutional neural network (CNN) for molecular fingerprint prediction based on data acquired by MS/MS. We used more than 680,000 MS/MS spectra obtained from the MoNA repository and NIST 20, representing about 36,000 compounds for training and testing our CNN model. The trained CNN model is implemented as a python package, MetFID. The package is available on GitHub for users to enter their MS/MS spectra and corresponding putative metabolite IDs to obtain ranked lists of metabolites. Better performance is achieved by MetFID in ranking putative metabolite IDs using the CASMI 2016 benchmark dataset compared to two other machine learning-based tools (CSI:FingerID and ChemDistiller).

Machine Learning-Based Earthquake Catalog and Tomography Characterize the Middle-Northern Section of the Xiaojiang Fault Zone

Abstract The left-lateral strike-slip Xiaojiang fault is an important tectonic boundary between Sichuan–Yunnan diamond block and the Yangtze block, which accommodated several M &amp;gt; 7.0 damaging earthquakes in the past ∼500 yr, as well as intense tectonic deformation. However, the seismogenesis of its middle-northern section are not well understood due to the lack of dense stations. In this study, we analyze one year of continuous seismic records from November 2019 to November 2020, which are recorded at a recently deployed dense seismic array. We build a high-precision earthquake catalog for the region using our recently developed machine learning-based earthquake location workflow (LOC-FLOW), which consists of machine learning phase picking, phase association, velocity model updating, and station correction, absolute location, and double-difference relative location. We then adopt a double-difference tomography method (tomoDD) to refine locations of 16,000 events and build a high-resolution 3D velocity model using both machine learning differential times and cross-correlation differential times. The seismicity distribution not only delineates detailed geometry of the main fault system but also characterizes several branch faults, including two echelon subfaults crossing the north–south-striking main fault. The velocity model shows strong lateral heterogeneities and exhibits a clear relationship to the seismicity distribution: the boundary of high- and low-velocity regions or high-velocity regions above low-velocity bodies accommodate the majority of earthquakes. The variation of the constructed 3D velocity model can be well explained by geological and tectonic settings of the region. In addition, we identify two seismic gaps, which accumulate stress and imply the potential of hosting future moderate-to-large earthquakes. Our study demonstrates, with the aid of LOC-FLOW and tomoDD, machine learning-based phase picks lead to promising performance in constraining high-precision earthquake catalogs and constructing high-resolution velocity models. Machine learning-based tools are becoming the next generation of routine earthquake analysis.

Innovative framework for the recognition of human activity in smart healthcare

Health services are one of the most difficult aspects of the massive influx of people into urban centers. With global digitization, especially in healthcare, the ability to collect, receive and present data has become a top priority. Currently, many applications in the healthcare business use Machine Learning to give individualized therapies, which will be more dynamic and efficient once personal health and predictive analytics are combined. Furthermore, machine learning-based tools aid in the treatment of patients starting at the ground level, with clinical practice diagnosis and suggestions. Massive amounts of data may be collected from smart devices in our digital age. Human activity recognition is a classification problem in which data is used to identify events performed by humans. This data is categorized and analyzed to discover patterns in the data that may be used to create future predictions. In theory, activity detection can provide significant societal benefits. Activity recognition and pattern discovery are two aspects of human activity comprehension. The first is concerned with detecting human activities accurately using a specified activity model.

Intelligent Low Complexity Resource Allocation Method for Integrated Satellite-Terrestrial Systems

Efficient frequency reusing schemes with multibeam satellite systems can be considered as a promising means of providing high-quality services with expensive and limited satellite bandwidth resources. This letter considers integrated satellite and terrestrial systems, where the same frequency bands are reused in the satellite beams and terrestrial cells. In such a condition, we propose an efficient power and frequency resource allocation scheme to minimize inter-component interference while the maximum throughput is provided to the users. The proposed scheme consists of two simple linear machine learning-based tools and a linear transformation tool. The simulation results show that the proposed scheme approximates the exhaustive search with a highly reduced complexity.

An efficient machine learning-based COVID-19 identification utilizing chest X-ray images

There is no well-known vaccine for coronavirus disease (COVID-19) with 100% efficiency. COVID-19 patients suffer from a lung infection, where lung-related problems can be effectively diagnosed with image techniques. The golden test for COVID-19 diagnosis is the RT-PCR test, which is costly, time-consuming and unavailable for various countries. Thus, machine learning-based tools are a viable solution. Here, we used a labelled chest X-ray of three categories, then performed data cleaning and augmentation to use the data in deep learning-based convolutional neural network (CNN) models. We compared the performance of different models that we gradually built and analyzed their accuracy. For that, we used 2905 chest X-ray scan samples. We were able to develop a model with the best accuracy of 97.44% for identifying COVID-19 using X-ray images. Thus, in this paper, we attested the feasibility of efficiently applying machine learning (ML) based models for medical image classification.