BioMedInformatics, Volume 2, Issue 4 (December 2022) – 17 articles

In recent years, the advent of new experimental methodologies for studying the high complexity of the human genome and proteome has led to the generation of an increasing amount of digital information, hence bioinformatics, which harnesses computer science, biology, and chemistry, playing a mandatory role for the analysis of the produced datasets. The emerging technology of Artificial Intelligence (AI), including Machine Learning (ML) and Artificial Neural Networks (ANNs), is nowadays at the core of biomedical research and has already paved the way for significant breakthroughs in both biological and medical sciences. AI and computer science have transformed traditional medicine into modern biomedicine, thus promising a new era in systems biology that will enhance drug discovery strategies and facilitate clinical practice. The current review defines the main categories of AI and thoroughly describes the fundamental principles of the widely used ML, ANNs and DL approaches. Furthermore, we aim to underline the determinant role of AI-based methods in various biological research fields, such as proteomics and drug design techniques, and finally, investigate the implication of AI in everyday clinical practice and healthcare systems. Finally, this review also highlights the challenges and future directions of AI in Modern Biomedical study. Full article

Behavioral factors can affect the blood glucose (BG) levels in people with type 1 diabetes (T1D), therefore, their effects need to be incorporated in blood glucose management for these individuals. Accordingly, in this work, we study the effect of two behavioral states, physical activity (PA) and stress state (SS), on BG fluctuations in individuals with T1D. We provide two methods for quantifying biomarkers related to PA and SS using raw acceleration (ACC) and electrodermal activity (EDA) data collected with a wearable device. We evaluate the impact of PA and SS on BG fluctuation by adding the derived behavior-related biomarkers in two cutting-edge deep learning-based glucose predictive models, a long short-term memory (LSTM) and a convolutional neural network (CNN)-LSTM network, for prediction horizons (PHs) of 30, 60 and 90 min. Through an ablation study, we demonstrate that incorporating the estimated behavior-related biomarkers improves the BG predictive model’s performance obtaining mean absolute error (MAE) 9.13 ± 0.95, 17.75 ± 1.93 and 31.85 ± 2.88 in [mg/dL], root mean square error (RMSE), 12.35 ± 1.06, 24.71 ± 2.31 and 41.64 ± 4.12 in [mg/dL], and coefficient of determination (R2), 95.34 ± 3.34, 78.87 ± 4.35 and 60.11 ± 4.76 in [%], for the LSTM model; and MAE 9.37 ± 0.88, 17.87 ± 1.67 and 29.47 ± 2.13 in [mg/dL], RMSE 12.51 ± 1.40, 24.37 ± 2.49 and 39.52 ± 3.89 in [mg/dL], and R2 94.65 ± 3.90, 78.37 ± 4.11 and 61.12 ± 4.30 in [%], for the CNN-LSTM model, respectively, across all PHs. Additionally, we illustrate the generalizability of the proposed models by performing both population- and patient-wise. Full article

Feature selection is a common step in data preprocessing that precedes machine learning to reduce data space and the computational cost of processing or obtaining the data. Filtering out uninformative variables is also important for knowledge discovery. By reducing the data space to only those components that are informative to the class structure, feature selection can simplify models so that they can be more easily interpreted by researchers in the field, reminiscent of explainable artificial intelligence. Knowledge discovery in complex data thus benefits from feature selection that aims to understand feature sets in the thematic context from which the data set originates. However, a single variable selected from a very small number of variables that are technically sufficient for AI training may make little immediate thematic sense, whereas the additional consideration of a variable discarded during feature selection could make scientific discovery very explicit. In this report, we propose an approach to explainable feature selection (XFS) based on a systematic reconsideration of unselected features. The difference between the respective classifications when training the algorithms with the selected features or with the unselected features provides a valid estimate of whether the relevant features in a data set have been selected and uninformative or trivial information was filtered out. It is shown that revisiting originally unselected variables in multivariate data sets allows for the detection of pathologies and errors in the feature selection that occasionally resulted in the failure to identify the most appropriate variables. Full article

Photorefractive keratectomy (PRK) has long been the method of choice for refractive surgery in pilots, and was FDA approved for U.S. Air Force aviators in 2000. We retrospectively reviewed the medical records of 16 male combat pilots (mean age 25.0 ± 5.5 years) who had undergone bilateral laser refractive surgery with surface ablation (alcohol-assisted PRK: 81.25%, transepithelial-PRK: 18.75%), and who had a mean baseline spherical equivalent (SE) of −2.1 ± 0.7 D in the right eye, and −2.0 ± 0.7 D in the left. The mean follow-up was 8.4 ± 6.6 months. On the last visit, the uncorrected visual acuity (UCVA) had improved from 0.75 ± 0.33 logMar to −0.02 ± 0.03 logMar (p < 0.001), and from 0.72 ± 0.36 logMar to −0.02 ± 0.05 logMar (p < 0.001), for the right and left eyes, respectively. The percentages of participants with a right eye UCVA of at least 0.0, −0.08, and −0.18 logMAR (6/6, 6/5, and 6/4 Snellen in meters) were 100%, 37.5%, and 6.2%, respectively, and for the left eye, 93.7%, 43.75%, and 6.2%, respectively. No complications occurred. This is the first study to assess refractive surgery outcomes in a cohort of Israeli combat pilots. Surface refractive surgery effectively improved UCVA and reduced spectacle reliance for the members of this visually demanding profession. Full article

Breast carcinoma (BCa) remains the second most common cause of cancer-related death among American women. Whereas estrogen receptor (ER) expression is typically regarded as a favorable prognostic indicator, a significant proportion of ER(+) patients still experience either de novo or acquired endocrine resistance. Previously, we have shown that the loss of orphan nuclear receptor NURR1 expression is associated with neoplastic transformation of the breast epithelium and shorter relapse-free survival (RFS) among systemically treated breast cancer (BCa) patients. Here, we further ascertain the prognostic value of NURR1 in BCa, and its differential expression among Black and White female BCa patients. We assessed the expression of NURR1 mRNA in BCa patients using the Cancer Genome Atlas (TGCA) and compared the occurrence of basal-like cancer and luminal A breast cancer subtypes. Expression levels were further stratified according to racial identity of the patient. We next assessed the correlation of NURR1 expression with Oncotype DX prognostic markers, and the association of NURR1 expression with relapse free survival in patients treated with endocrine therapy. Our study shows that NURR1 mRNA expression is differentially correlated with luminal A vs. basal-like cancer BCa and is predictive of poor relapse-free survival, confirming a similar trend observed in our previous studies using microarray data. NURR1 expression was positively correlated with expression of Oncotype DX biomarkers associated with estrogen responsiveness, while being inversely correlated with biomarkers associated with cell proliferation. Furthermore, we observed that NURR1 expression was positively associated with greater relapse-free survival at 5 years among patients treated with endocrine therapy. Interestingly, we found that among Black women with luminal A BCa, NURR1 expression was repressed in comparison to White women with the same subtype. Full article

Individualized, serial production of innovative implants is a major area of application for additive manufacturing in the field of medicine. Individualized healthcare requires faster delivery of the implant to the clinic or hospital facility. The total manufacturing process, including data generation using 3D drawings, imaging techniques, 3D printing and post-processing, usually takes up to a week, especially implants from risk class III, which requires qualified equipment and a validated process. In this study, we describe how to develop a new biomechanical model for dental implants from its conception for the patent to the final product which is ready to be manufactured using additive manufacturing. The benefits and limitations of titanium metal printing for dental implant prototypes are presented by the authors. Full article

Accurate and early detection of causes of pneumonia is important for implementing fast treatment and preventive strategies, reducing the burden of infections, and establishing more effective ways of interventions. After the outbreak of COVID-19, the new cases of pneumonia and conditions of breathing problems called acute respiratory distress syndrome have increased. Chest radiography, known as CXR or simply X-ray has become a significant source to diagnose COVID-19-infected pneumonia in designated institutions and hospitals. It is essential to develop automated computer systems to assist doctors and medical experts to diagnose pneumonia in a fast and reliable manner. In this work, we propose a deep learning (DL)-based computer-aided diagnosis system for rapid and easy detection of pneumonia using X-ray images. To improve classification accuracy and faster conversion of the models, we employ transfer learning and parallel computing techniques using well-known DL models such as VGG19 and ResNet50. Experiments are conducted on the large COVID-QU-Ex dataset of X-ray images with three classes, such as COVID-19-infected pneumonia, non-COVID-19 infections (other viral and bacterial pneumonia), and normal (uninfected) images. The proposed model outperformed compared methodologies, achieving an average classification accuracy of 96.6%. Experimental results demonstrate that the proposed method is effective in diagnosing pneumonia using X-ray images. Full article

Mitochondrial DNA (mtDNA) damage is closely associated with typical diseases of aging, such as Alzheimer’s or Parkinson’s disease, and other health conditions, such as infertility. This damage manifests in reduced mitochondrial copy number and deletion mutations in mtDNA. Consequently, the analysis of mitochondrial damage by determining the parameters copy number and deletion ratio using quantitative real-time PCR (qPCR) is of interest for clinical diagnostics. To bring the findings from research into laboratory practice, a suitable and reliable process is needed, which must be thoroughly validated. This process includes the software used for the analysis, which must meet extensive regulatory and process requirements. Existing software does not adequately implement the requirements of laboratories and, in particular, does not provide direct support for the calculation of the aforementioned mtDNA parameters. The paper discusses the development of a new software-based analysis workflow that is designed specifically for laboratories to help with the calculation of mtDNA parameters. The software was developed using the User-Centered Design method and is based on the recently introduced prototype, “PlateFlow”. Initial user tests provide positive feedback. In the future, this workflow could form the basis for validations of mitochondrial tests in medical laboratories. Full article

Background: High-quality oligonucleotides for molecular amplification and detection procedures of diverse target sequences depend on sequence homology. Processing input sequences and identifying homogeneous regions in alignments can be carried out by hand only if they are small and contain sequences of high similarity. Finding the best regions for large and inhomogeneous alignments needs to be automated. Results: The ConsensusPrime pipeline was developed to sort out redundant and technical interfering data in multiple sequence alignments and detect the most homologous regions from multiple sequences. It automates the prediction of optimal consensus primers for molecular analytical and sequence-based procedures/assays. Conclusion: ConsensusPrime is a fast and easy-to-use pipeline for predicting optimal consensus primers that is executable on local systems without depending on external resources and web services. An implementation in a Docker image ensures platform-independent executability and installability despite the combination of multiple programs. The source code and installation instructions are publicly available on GitHub. Full article

The retinal degenerative disease retinitis pigmentosa (RP) is a genetic disease that is the most common cause of blindness in adults. In 2016, Chow et. al. identified over 100 candidate modifier genes for RP through the genome-wide analysis of 173 inbred strains from the Drosophila Genetic Reference Panel (DGRP). However, this type of analysis may miss some modifiers lying in trans to the variation. In this paper, we propose an alternative approach to identify transcripts whose expression is significantly altered in strains demonstrating extreme phenotypes. The differences in the eye size phenotype will, therefore, be associated directly with changes in gene expression rather than indirectly through genetic variation that might then be linked to changes in gene expression. Gene expression data are obtained from the DGRP2 database, where each strain is represented by up to two replicates. The proposed algorithmic approach first chooses the strains’ replicate combination that best represents the relationship between gene expression level and eye size. The extensive correlation analysis identified several genes with known relationships to eye development, along with another set of genes with unknown functions in eye development. The modifiers identified in this analysis can be validated and characterized in biological systems. Full article

The adoption of “artificial intelligence (AI) in drug discovery”, where AI is used in the process of pharmaceutical research and development, is progressing. By using the ability to process large amounts of data, which is a characteristic of AI, and achieving advanced data analysis and inference, there are benefits such as shortening development time, reducing costs, and reducing the workload of researchers. There are various problems in drug development, but the following two issues are particularly problematic: (1) the yearly increases in development time and cost of drugs and (2) the difficulty in finding highly accurate target genes. Therefore, screening and simulation using AI are expected. Researchers have high demands for data collection and the utilization of infrastructure for AI analysis. In the field of drug discovery, for example, interest in data use increases with the amount of chemical or biological data available. The application of AI in drug discovery is becoming more active due to improvement in computer processing power and the development and spread of machine-learning frameworks, including deep learning. To evaluate performance, various statistical indices have been introduced. However, the factors affected in performance have not been revealed completely. In this study, we summarized and reviewed the applications of deep learning for drug discovery with BigData. Full article

Determining whether emergency catheterization is necessary for treating heart disease in older patients in rural hospitals is important. Their transportation may be affected by ageism. This retrospective cohort study investigated the relationship between patient factors and emergency catheterization in rural hospitals in patients >65 years old who visited the emergency department and were transferred to tertiary hospitals. Factors related to emergency catheterization were analyzed using a logistic regression model. The average age of the exposure and control groups was 77.61 (standard deviation [SD], 13.76) and 74.90 (SD, 16.18) years, respectively. Men accounted for 54.8 and 67.5% of patients in the exposure and control groups, respectively. Factors related to emergency catheterization were Charlson comorbidity index ≥5 (odds ratio [OR], 0.23; 95% confidence interval [CI], 0.06–0.94) and electrocardiogram (ECG) changes (OR, 3.24; 95% CI, 1.00–10.50). In these patients, age, time from onset to transfer, and serum troponin level were not significantly related to emergency catheterization, while ECG changes correlated with the indication for emergency catheterization. Emergency catheterization patients did not confirm that ageism was present. The decision for transfer to tertiary hospitals should consider comorbidities and ECG changes and should not be influenced by age, onset, and troponin level. Full article

Immune checkpoint blockade targeting PDCD1 (PD-1) or CD274 (PD-L1) has demonstrated efficacy and interest across multiple cancers. However, the exact determinants of the response and cancer-specific molecular features remain unclear. A recent pan-cancer study identified a PDCD1/CD274-related immunotherapy network of 40 genes that had differential patient survival associations across multiple cancers. However, the survival relevance of this network in mesothelioma could not be assessed due to a lack of available survival data for the mesothelioma study included. Mesothelioma, a rare cancer that most commonly arises in the pleural membranes around the lung, does have immune checkpoint blockade as an approved treatment strategy, yet questions over its efficacy remain. RNA-seq data from 87 pleural mesothelioma patients were interrogated on cBioPortal to assess the role of the PDCD1/CD274 network identified in a previous study, in addition to identifying repurposed drugs that may have therapeutic efficacy. Extensive literature searches were conducted to identify known information from the literature around the genes shown to impact patient survival (CCR5, GATD3A/GATD3, CXCR6, GZMA, and TBC1D10C). The same literature validation was performed for putative repurposed drugs that were identified as potential immunotherapeutic adjuvants in the context of mesothelioma (disulfiram, terfenadine, maraviroc, clioquinol, chloroxine, and oxyphenbutazone). Only disulfiram returned a specifically focused research article based on the literature search. This article demonstrated cytotoxicity in a panel of five human MPM cell lines of mixed histology (epithelioid, biphasic, and sarcomatoid). There was little information on the remaining five drugs, yet the clear preclinical efficacy of disulfiram validates the methodology used herein and prompts further exploration of the remaining drugs in mesothelioma. This study ultimately sheds light on novel preclinical information of genes related to PDCD1/CD274 in mesothelioma, as well as identifying putative drugs that may have therapeutic efficacy either independently or as an immunotherapeutic adjuvant. Full article

Bacteria that are capable of organizing themselves as biofilms are an important public health issue. Knowledge discovery focusing on the ability to swarm and conquer the surroundings to form persistent colonies is therefore very important for microbiological research communities that focus on a clinical perspective. Here, we demonstrate how a machine learning workflow can be used to create useful models that are capable of discriminating distinct associated growth behaviors along distinct phenotypes. Based on basic gray-scale images, we provide a processing pipeline for binary image generation, making the workflow accessible for imaging data from a wide range of devices and conditions. The workflow includes a locally estimated regression model that easily applies to growth-related data and a shape analysis using identified principal components. Finally, we apply a density-based clustering application with noise (DBSCAN) to extract and analyze characteristic, general features explained by colony shapes and areas to discriminate distinct Bacillus subtilis phenotypes. Our results suggest that the differences regarding their ability to swarm and subsequently conquer the medium that surrounds them result in characteristic features. The differences along the time scales of the distinct latency for the colony formation give insights into the ability to invade the surroundings and therefore could serve as a useful monitoring tool. Full article

Omicron is the dominant strain of COVID-19 in the United States and worldwide. Although this variant is highly transmissible and may evade natural immunity, vaccines, and therapeutic antibodies, preclinical results in animal models and clinical data in humans suggest omicron causes a less severe form of infection. The molecular basis for the attenuation of virulence when compared to previous variants is currently not well understood. Using protein–ligand docking simulations to evaluate and compare the capacity of SARS-CoV-2 spike-1 proteins with the different COVID-19 variants to bind to the human α7nAChr (i.e., the core receptor under the control of the vagus nerve regulating the parasympathetic nervous system and the cholinergic anti-inflammatory pathway), we found that 10 out of the 14 mutated residues on the RBD of the B.1.1.529 (Omicron) spike, compared to between 0 and 2 in all previous variants, were present at the interaction interface of the α7nAChr. We also demonstrated, through protein–protein docking simulations, that these genetic alterations cause a dramatic decrease in the ability of the Omicron SARS-CoV-2 spike-1 protein to bind to the α7nAChr. These results suggest, for the first time, that the attenuated nature of Omicron infection in humans and animals compared to previous variants may be attributable to a particular set of genetic alterations, specifically affecting the binding site of the SARS-CoV-2 spike-1 protein to the α7nAChr. Full article

Knowledge discovery in biomedical data using supervised methods assumes that the data contain structure relevant to the class structure if a classifier can be trained to assign a case to the correct class better than by guessing. In this setting, acceptance or rejection of a scientific hypothesis may depend critically on the ability to classify cases better than randomly, without high classification performance being the primary goal. Random forests are often chosen for knowledge-discovery tasks because they are considered a powerful classifier that does not require sophisticated data transformation or hyperparameter tuning and can be regarded as a reference classifier for tabular numerical data. Here, we report a case where the failure of random forests using the default hyperparameter settings in the standard implementations of R and Python would have led to the rejection of the hypothesis that the data contained structure relevant to the class structure. After tuning the hyperparameters, classification performance increased from 56% to 65% balanced accuracy in R, and from 55% to 67% balanced accuracy in Python. More importantly, the 95% confidence intervals in the tuned versions were to the right of the value of 50% that characterizes guessing-level classification. Thus, tuning provided the desired evidence that the data structure supported the class structure of the data set. In this case, the tuning made more than a quantitative difference in the form of slightly better classification accuracy, but significantly changed the interpretation of the data set. This is especially true when classification performance is low and a small improvement increases the balanced accuracy to over 50% when guessing. Full article

In this paper, a novel approach for abnormal gait and tremor detection using a smart walking cane is introduced. Periodic muscle movement associated with Parkinson’s disease, such as arm shaking, vibrating arm, trembling fingers, rhythmic wrist movements, normal and abnormal walking pattern, was learned and classified with linear discriminant analysis. Although detecting symptoms related to disease with walking sticks might look trivial at first, throughout history, a cane or walking stick has been used as an assistive device to aid in ambulating, especially in the elderly and disabled, so embedding smart devices (that can learn ambulating pattern and detect anomalies associated with it) in the cane will help in early detection of diseases and facilitate early intervention. This approach is non-intrusive, and privacy issues being experienced in visual models do not arise, as users do not need to wear any special bracelet or wrist monitoring, and they only need to pick up the cane when they wish to move. The simplicity and efficient usage of a technique for detecting ambulatory anomalies is also demonstrated in this research. We extracted step counts, fall data and other valuable features from the cane, and detected anomalies by using isolation forest and one-class support vector machine (SVM) methods. Falls were detected easily and naturally with the cane, which had different alert modes (a soft alert when the cane lost equilibrium and was picked up within 15 s, and a strong alert otherwise). Intervention systems are proposed to forestall and limit the possibility of a type 2 error. Full article