ARTICLE | doi:10.20944/preprints202307.0355.v1
Subject: Medicine And Pharmacology, Otolaryngology Keywords: Third window; X-linked gusher; hearing loss; audiovestibular; POU3F4
Online: 6 July 2023 (09:16:03 CEST)
Conductive hearing losses are typically present in disorders of the external/middle ear. However, there is a rare group of inner ear conditions called third windows that can also generate a conductive hearing loss. This is due to an abnormal connection between the middle and the inner ear or between the inner ear and the cranial cavity. X-linked gusher disorder is an extremely rare congenital inner ear dysplastic syndrome with such an abnormal connection due to a characteristic incomplete cochlear partition type 3 and an incomplete internal auditory meatus fundus. The disorder is inherited in an X-linked fashion due to the mutation of the POU3F4 gene. We present 2 siblings diagnosed with the condition and their long term follow ups. They both presented with audiovestibular symptoms and showed progressive mixed losses and bilateral vestibular weakness. They were treated with cochlear implant, digital amplification and with vestibular rehabilitation. Significant others around them were involved in their journey with the medical team and in both, a very favourable outcome was achieved. This is the first time that we are reporting evolving audiovestibular function with vestibular quantification in X-linked gusher disorder and emphasize on the multidisciplinary holistic approach to manage these children effectively.
ARTICLE | doi:10.20944/preprints202307.0529.v1
Subject: Computer Science And Mathematics, Artificial Intelligence And Machine Learning Keywords: machine learning; accuracy; complexity; entropy; landslide susceptibility mapping; dimensionality reduction; Principal Component Analysis (PCA)
Online: 10 July 2023 (11:11:27 CEST)
In this study, our primary objective was to analyze the tradeoff between accuracy and complexity in machine learning models, with a specific focus on the impact of reducing complexity and entropy on the production of landslide susceptibility maps. We aimed to investigate how simplifying the model and reducing entropy can affect the capture of complex patterns in the susceptibility maps. To achieve this, we conducted a comprehensive evaluation of various machine-learning algorithms for classification tasks. We compared the performance of these algorithms in terms of accuracy and complexity, considering both "before" and "after" scenarios of dimensionality reduction using Principal Component Analysis (PCA). Our findings revealed that reducing complexity and lowering entropy can lead to an increase in model accuracy. However, we also observed that this reduction in complexity comes at the cost of losing important complex patterns in the produced landslide susceptibility maps. By simplifying the model and reducing entropy, certain intricate relationships and uncertain patterns may be overlooked, resulting in a loss of information and potentially compromising the accuracy of the susceptibility maps. The analysis encompassed a diverse range of machine learning algorithms, including Random Forest (RF), Extra Trees (EXT), XGboost, LightGBM, Catboost, Naive Bayes (NB), K-Nearest Neighbors (KNN), Gradient Boosting Machine (GBM), and Decision Trees (DT). Each algorithm was evaluated for its strengths and limitations, considering the tradeoff between accuracy and complexity. Before dimensionality reduction, the algorithms demonstrated promising results, with RF exhibiting excellent AUC/ROC scores and average accuracy. However, computational costs were noted as a potential drawback for RF, especially when dealing with large datasets. EXT showcased robust performance and good accuracy, while XGboost demonstrated its ability to handle complex relationships within large datasets, albeit requiring careful hyperparameter tuning. The efficiency and scalability of LightGBM made it a suitable choice for large datasets, although it displayed sensitivity to class imbalance. Catboost excelled in handling categorical features, but longer training times were observed for larger datasets. NB showcased simplicity and computational efficiency but assumed independence among features. KNN, known for its capability to capture local patterns and spatial relationships, was found to be sensitive to the choice of distance metric. GBM, while capturing complex relationships effectively, was prone to overfitting without proper regularization. DT, with its interpretability and ease of understanding, faced limitations in terms of overfitting and limited generalization. After dimensionality reduction, certain algorithms exhibited improvements in their AUC/ROC scores and average accuracy, including RF, EXT, XGboost, and LightGBM. However, for a few algorithms, such as NB and DT, a decrease in performance was observed. This study provides valuable insights into the performance characteristics, strengths, and limitations of various machine learning algorithms in classification tasks. Researchers and practitioners can utilize these findings to make informed decisions when selecting algorithms for their specific datasets and requirements. We also aim to identify the potential factors contributing to the high accuracy rates obtained from these ensembled algorithms and explore possible shortcomings of non-ensembled algorithms that may result in lower accuracy rates. By conducting a comprehensive analysis of these algorithms, we seek to provide valuable insights into the benefits and limitations of ensembled approaches for landslide susceptibility mapping. Our study sheds light on the challenges faced when balancing accuracy and complexity in machine learning models for landslide susceptibility mapping. It emphasizes the importance of carefully considering the level of complexity and entropy reduction in relation to the specific patterns and uncertainties present in the data. By providing insights into this tradeoff, our research aims to assist researchers and practitioners in making informed decisions regarding model complexity and entropy reduction, ultimately improving the quality and interpretability of landslide susceptibility maps.
ARTICLE | doi:10.20944/preprints202212.0211.v1
Subject: Biology And Life Sciences, Biochemistry And Molecular Biology Keywords: klotho; estrogen; hippocampus; chronic stress; sex difference; stress resilience
Online: 13 December 2022 (01:09:36 CET)
Klotho (KL) is a glycosyl hydrolase and aging-suppressor gene. Stress is a risk factor for depression and anxiety that are highly comorbid with each other. The aim of this study was to determine KL is regulated by estrogen and plays an important role in sex differences in stress resilience. Our results showed that KL was regulated by estrogen in rat hippocampal neurons in vivo and in vitro and was essential for estrogen-mediated increase in the number of presynaptic vesicular glutamate transporter 1 (Vglut1) positive clusters on the dendrites of hippocampal neurons. The role of KL in sex differences in stress responses was examined in rats using three-week chronic unpredictable mild stress (CUMS). CUMS produced a deficit in spatial learning and memory, anhedonic-like and anxiety-like behaviors in male but not female rats, which was accompanied by a reduction in KL protein levels in the hippocampus of male, but not female rats. This demonstrated the resilience of female rats to CUMS. Interestingly, knockdown of KL protein levels in the rat hippocampus of both sexes caused a decrease in stress resilience in both sexes, especially in female rats. These results suggest that regulation of KL by estrogen plays an important role in estrogen-mediated synapse formation, and KL plays a critical role in the sex differences in cognitive deficit, anhedonic-like and anxiety-like behaviors induced by chronic stress in rats, highlighting an important role of KL in sex differences in stress resilience.
ARTICLE | doi:10.20944/preprints202306.1784.v1
Subject: Computer Science And Mathematics, Computer Science Keywords: machine learning algorithms; hyperparameters; hyperparameter optimization; spatial data; Bayesian optimization; metaheuristic algorithms
Online: 26 June 2023 (10:10:33 CEST)
Algorithms for machine learning have found extensive use in numerous fields and applications. One important aspect of effectively utilizing these algorithms is tuning the hyperparameters to match the specific task at hand. The selection and configuration of hyperparameters directly impact the performance of machine learning models. Achieving optimal hyperparameter settings often requires a deep understanding of the underlying models and the appropriate optimization techniques. While there are many automatic optimization techniques available, each with its own advantages and disadvantages, this article focuses on hyperparameter optimization for well-known machine learning models. It explores cutting-edge optimization methods and provides guidance on applying them to different machine learning algorithms. The article also presents real-world applications of hyperparameter optimization by conducting tests on spatial data collections for landslide susceptibility mapping. Based on the experiment's results, both Bayesian optimization and metaheuristic algorithms showed promising performance compared to baseline algorithms. For example, the metaheuristic algorithm improved the overall accuracy of the random forest model. Additionally, Bayesian algorithms, such as Gaussian processes, performed well for models like KNN and SVM. The paper thoroughly discusses the reasons behind the efficiency of these algorithms. By successfully identifying appropriate hyperparameter configurations, this research paper aims to assist researchers, spatial data analysts, and industrial users in developing machine learning models more effectively. The findings and insights provided in this paper can contribute to enhancing the performance and applicability of machine learning algorithms in various domains.
ARTICLE | doi:10.20944/preprints202307.1467.v1
Subject: Computer Science And Mathematics, Artificial Intelligence And Machine Learning Keywords: artificial neural networks; Bayesian techniques; metaheuristic techniques; hyperparameters; feature selection techniques
Online: 26 July 2023 (03:37:57 CEST)
The most frequent, noticeable, and frequent natural calamity in the karakoram region is landslides. Extreme landslides have occurred frequently along Karakoram highway, particularly during the monsoon, causing a major loss of life and property. Therefore, it was necessary to look for a solution to increase growth and vigilance in order to lessen losses related to landslides caused by natural disasters. By utilizing contemporary technologies, an early warning system might be developed. Artificial neural networks (ANNs) are widely used nowadays across many industries. This paper's major goal is to provide new integrative models for assessing landslide susceptibility in a prone area of north of Pakistan. To do this, the training of an artificial neural network (ANN) is supervised using metaheuristic and Bayesian techniques: particle swarm optimization algorithm (PSO), Genetic algorithm (GA), Bayesian optimization Gaussian process (BO_GP), and Bayesian optimization Gaussian process (BO_TPE). 304 previous landslides and the eight most prevalent conditioning elements combine to form a geographical database. The models are hyper-parameter optimized, and the best ones are employed to generate the susceptibility maps. The area under the receiving operating characteristic curve (AUROC) accuracy index found demonstrated that the maps produced by both Bayesian and metaheuristic algorithms are highly accurate. The effectiveness and efficiency of applying artificial neural networks (ANNs) for landslide mapping, susceptibility analysis, and forecasting are studied in this research it’s observed from experimentation that the performance differences for GA, BO_GP, and PSO compared to BO_TPE are relatively small, ranging from 0.3166% to 1.8399%. This suggests that these techniques achieved comparable performance to BO_TPE in terms of AUC. However, it's important to note that the significance of these differences can vary depending on the specific context and requirements of the ML task. Additionally in this study, we explore eight feature selection algorithms to determine the geospatial variable importance for landslide susceptibility mapping along the KKH. The algorithms considered include Information Gain, Gain Ratio, OneR Classifier, Subset Evaluators, Principal Components, Relief Attribute Evaluator, Correlation, and Symmetrical Uncertainty. These algorithms enable us to evaluate the relevance and significance of different geospatial variables in predicting landslide susceptibility. By applying these feature selection algorithms, we aim to identify the most influential geospatial variables that contribute to landslide occurrences along the KKH. The algorithms encompass a diverse range of techniques, such as measuring entropy reduction, accounting for attribute bias, generating single rules, evaluating feature subsets, reducing dimensionality, and assessing correlation and information sharing. The findings of this study will provide valuable insights into the critical geospatial variables associated with landslide susceptibility along the KKH. These insights can aid in the development of effective landslide mitigation strategies, infrastructure planning, and targeted hazard management efforts. Additionally, the study contributes to the field of geospatial analysis by showcasing the applicability and effectiveness of various feature selection algorithms in the context of landslide susceptibility mapping.
ARTICLE | doi:10.20944/preprints202309.1723.v1
Subject: Biology And Life Sciences, Biology And Biotechnology Keywords: Mango tree; SSR markers; mango varieties; genetic diversity; cultivar identification diagram, loci, polymorphism; genetic resources in Pakistan; breeding; cultivar development
Online: 26 September 2023 (03:48:00 CEST)
Assessment of the genetic distinctiveness of a cultivar through morphological descriptors is an important tool for both the registration and the protection. New mango genotypes have been improved using valuable diverse germplasm resources to ensure food security. DNA fingerprinting based simple sequence repeats (SSR)-markers have been the most broadly used, effective and accurate in evaluation of genetic characterization of a cultivar. Molecular breeding is an effective source of genetic gain after improvement of fruit trees using marker assisted genomic selection. Mango (Mangifera indica L.) is an allotetraploid (2n = 4X= 40) drupe fruit and has high nutritional value belongs to genus Mangifera and family Anacardiaceae. Mango cultivars are used with worldwide acceptance to pharmacological, ethnomedical, and phytochemical industries. This study investigated the molecular evaluation of a new mango cultivar ‘Azeem Chaunsa’ using a set of the most effective 50 hyper-variable polymorphic SSR markers. Highly specific DNA fingerprints were identified in the genome of this mango cultivar, ‘Azeem Chaunsa’ compared with three standard cultivars such as Sindhri, Samar Bahisht (S.B) Chaunsa and Sufaid Chaunsa. Our results showed that SSR markers could efficiently assess genetic diversity in mango. An agglomerative hierarchical clustering method was used to construct dendrogram based on the ‘Unweighted Pair- Group Method with Arithmetic Mean’ (UPGMA). The genetic similarity coefficients were recorded between the mangos cultivars ranged from 0.49 to 0.67. Cultivar identification (CID) evaluates association among standard cultivars and Azeem Chaunsa and further concludes significant variations. CID results concluded that cultivar ‘Azeem Chaunsa’ varied significantly from the check cultivar, Sindhri (46.2%), S.B Chaunsa (45%) and Sufaid Chaunsa (46.7%). The results obtained in this study will orient cultivar identification strategies for a successful future mango breeding programmes in the context of climate change.
ARTICLE | doi:10.20944/preprints202310.1821.v1
Subject: Public Health And Healthcare, Public Health And Health Services Keywords: type-2 diabetes mellitus; pro-inflammatory chemokines; obesity; CCL1; CCL2; CCL4; CCL5
Online: 27 October 2023 (15:52:34 CEST)
Background. Type 2 diabetes mellitus (T2DM) is becoming a major global health concern, especially in poorer nations. The high prevalence of obesity and the ensuing diabetes is attributed to rapid economic progress, physical inactivity, consumption of high-calorie foods, and changing lifestyles. Objectives. We investigated the role of pro-inflammatory chemokines; CCL1, 2, 4 and 5 with varying levels of obesity in T2DM in the Asir region of Saudi Arabia. Materials and methods. 170 confirmed T2DM patients and a normal control group were enrolled. The demographic data, serum levels CCL-1, 2, 4 and 5 and the biochemical indices were assessed in patients and control groups by standard procedures. Results. T2DM patients were divided into four groups: A (normal body weight), B (Overweight), C (obese), and D (highly obese). n controls. We observed that male and female control patients had similar mean serum concentrations of pro-inflammatory chemokines CCL 1, 2, 4, and 5. Chemokines CCL1, CCL2, and CCL4 in the serum of Type 2 Diabetes Mellitus patients with normal or overweight body weights were significantly higher than the control group, regardless of gender. In T2DM individuals with obesity and severe obesity, the rise was most significant. There was a progressive rise in the concentrations of CCL1, 2, and 4 in T2DM) patients with increasing BMI. Serum CCL5 levels increased significantly in all T2DM patient groups. The increase in CCL5 was more predominant in normal-weight people as compared to overweight and obese T2DM patients. Conclusions. Male and female control patients had similar serum levels of pro-inflammatory chemokines CCL 1, 2, 4, and 5. The progressive rise in blood concentrations of three Pro-inflammatory chemokines; CCL1, 2, and 4 in T2DM patients with increasing BMI supports the idea that dyslipidemia and obesity contribute to chronic inflammation and insulin resistance. Serum CCL5 levels increased significantly in all T2DM patient groups. The selective and more pronounced increase of CCL5 in the T2DM group with normal BMI as compared to patients with varying degrees of obesity, was rather surprising. Further research is needed to determine if CCL5 under-expression in overweight and obese T2DM patients is due to some unexplained counterbalancing processes.
ARTICLE | doi:10.20944/preprints202310.1192.v1
Subject: Medicine And Pharmacology, Endocrinology And Metabolism Keywords: Diabetes; Pro and anti-inflammatory cytokines; T2DM in Saudi Arabia; Pathogenesis; IL-10; IL-19; and IL 22; Obesity
Online: 18 October 2023 (13:47:02 CEST)
Background. Type 2 diabetes mellitus (T2DM) is becoming a major global health concern, especially in poorer nations. The high prevalence of obesity and the ensuing diabetes I attributed to rapid economic progress, physical inactivity, consumption of high-calorie foods, and changing lifestyles. Objectives. We investigated the role of interleukins-10 , 19 and 22 with varying levels of obesity in T2DM in the Asir region of Saudi Arabia. Materials and methods. 170 confirmed T2DM patients and a control group were enrolled. The demographic data, serum levels of IL-10, IL-19, and IL-22, and biochemical indices were assessed in patients and control groups by standard procedures. Results. T2DM patients were divided into four groups: A (normal body weight), B (Overweight), C (obese), and D (highly obese). Both male and female T2DM subjects in Group A showed significant decreases in IL-10 levels as compared to controls; however, there were insignificant changes in IL-10 levels in Groups C and D. T2DM patients in groups C and D, in both males and females, depicted very significant (p 0.001) increases in IL-19 levels as compared to controls and group A. Patient groups A to D displayed a progressive elevation of Il-22 levels irrespective of gender, although significant alterations were seen only in groups B to D, with p 0.05 for group B and p 0.01 for groups C and D respectively, as compared to healthy controls. Conclusions. IL-10 showed a strong relationship with T2DM in males with varying degrees of obesity, but females depicted relatively higher IL-10 levels in obese and highly obese groups, pointing to a protective phenomenon. IL-19 levels showed significant increases in all four groups, irrespective of gender. IL-22 appears to be unrelated to T2DM per se but shows an association with varying degrees of obesity.