Prediction of Genetic Groups within Brettanomyces bruxellensis through Cell Morphology Using a Deep Learning Tool

It is essential to discriminate between B. bruxellensis isolates at the strain level, because stress resistance capacities are strain dependent and also related to the genetic groups (GG). In this work, we investigated further the correlation between genetic groups and cell polymorphism by analysing optical microscopy images via deep learning. A Convolutional Neural Network (CNN) was trained to discriminate between 74 different B. bruxellensis isolates belonging to 4 of the 6 genetic groups described. Compared to the microsatellite analysis, the CNN enabled the prediction of the genetic groups of B. bruxellensis isolates with 96.6 % accuracy in a faster and cheaper way and with the same genetic group affiliations. Based on these very promising results, further research is needed to validate this technique for all genetic groups.

Brettanomyces bruxellensis is the main wine spoiler yeast all over the world, yet the structure of the populations associated with winemaking remains elusive. In this work, we considered 1411 wine isolates from 21 countries that were genotyped using twelve microsatellite markers. We confirmed that B. bruxellensis isolates from wine environments show high genetic diversity, with 58 and 42% of putative triploid and diploid individuals respectively distributed in 5 main genetic groups. The distribution in the genetic groups varied greatly depending on the country and/or the wine-producing region. However, the two possible triploid wine groups showing sulfite resistance/tolerance were identified in almost all regions/countries. Genetically identical isolates were also identified. The analysis of these clone groups revealed that a given genotype could be isolated repeatedly in the same winery over decades, demonstrating unsuspected persistence ability. Besides cellar residency, a great geographic dispersal was also evidenced, with some genotypes isolated in wines from different continents. Finally, the study of old isolates and/or isolates from old vintages revealed that only the diploid groups were identified prior 1990 vintages. The putative triploid groups were identified in subsequent vintages, and their proportion has increased steadily these last decades, suggesting adaptation to winemaking practices such as sulfite use. A possible evolutionary scenario explaining these results is discussed.

Brettanomyces bruxellensis is the main wine spoiler yeast all over the world, yet the structure of the populations associated with winemaking remains elusive. In this work, we considered 1411 wine isolates from 21 countries that were genotyped using twelve microsatellite markers. We confirmed that B. bruxellensis isolates from wine environments show high genetic diversity, with 58 and 42% of putative triploid and diploid individuals respectively distributed in 5 main genetic groups. The distribution in the genetic groups varied greatly depending on the country and/or the wine-producing region. However, the two possible triploid wine groups showing sulfite resistance/tolerance were identified in almost all regions/countries. Genetically identical isolates were also identified. The analysis of these clone groups revealed that a given genotype could be isolated repeatedly in the same winery over decades, demonstrating unsuspected persistence ability. Besides cellar residency, a great geographic dispersal was also evidenced, with some genotypes isolated in wines from different continents. Finally, the study of old isolates and/or isolates from old vintages revealed that only the diploid groups were identified prior 1990 vintages. The putative triploid groups were identified in subsequent vintages, and their proportion has increased steadily these last decades, suggesting adaptation to winemaking practices such as sulfite use. A possible evolutionary scenario explaining these results is discussed.

To evaluate the coumarate descarboxylase (CD) and vinylphenol reductase (VR) activities in Dekkera bruxellensis isolates and study their relationship to the growth rate, protein profile and random amplified polymorphic DNA (RAPD) molecular pattern. CD and VR activities were quantified, as well, the growth rate, intracellular protein profile and molecular analysis (RAPD) were determined in 12 isolates of D. bruxellensis. All the isolates studied showed CD activity, but only some showed VR activity. Those isolates with the greatest growth rate did not present a different protein profile from the others. The FASC showed a relationship between RAPD molecular patterns and VR activity. CD activity is common to all of the D. bruxellensis isolates. This was not the case with VR activity, which was detected at a low percentage in the analysed micro-organisms. A correlation was observed between VR activity and the RAPD patterns. This is the first study that quantifies the CD and VR enzyme activities in D. bruxellensis, demonstrating that these activities are not present in all isolates of this yeast.

Application of Deep Learning Convolutional Neural Networks for Internal Tablet Defect Detection: High Accuracy, Throughput, and Adaptability

Dekkera bruxellensis is a spoilage yeast in wine and fuel ethanol fermentations able to produce volatile phenols from hydroxycinnamic acids by the action of the enzymes cinnamate decarboxylase (CD) and vinyphenol reductase (VR) in wine. However, there is no information about this ability in the bioethanol industry. This work evaluated CD and VR activities and 4-ethylphenol production from p-coumaric acid by three strains of D. bruxellensis and PE-2, an industrial Saccharomyces cerevisiae strain. Single and multiple-cycle batch fermentations in molasses and sugarcane juice were carried out. Dekkera bruxellensis strains showed similar CD activity but differences in VR activity. No production of 4-ethylphenol by S. cerevisiae in any fermentation system or media was observed. The concentrations of 4-ethylphenol peaked during active growth of D. bruxellensis in single-cycle fermentation but they were lower than in multiple-cycle fermentation. Higher concentrations were observed in molasses with molar conversion (p-coumaric acid to 4-ethylphenol) ranging from 45% to 85%. As the first report on 4-ethylphenol production in sugarcane musts by D. bruxellensis in industry-like conditions, it opens up a new avenue to investigate its effect on the viability and fermentative capacity of S. cerevisiae as well as to understand the interaction between the yeasts in the bioethanol industry.

Forty-three yeast isolates derived from various fermented foods, alcoholic beverages and traditional inocula of Western Himalayas were characterized by using traditional and molecular techniques. Traditional characterization identified these isolates as belonging to seven genera and eight species. Twenty-three yeast isolates were identified as Saccharomyces cerevisiae, six as Debaromyces hansenii, five as Issatchenkia orientalis, four as Saccharomycesfermentati, two as Schizosaccharomyces pombe and one each as Endomyces fibuliger, Brettanomyces bruxellensis and Candida tropicalis. The molecular characterization using four marker systems i.e. universal rice primers (URP), randomly amplified polymorphic DNA (RAPD), inter simple sequence repeat (ISSR) and delta typing was carried out, which revealed strainal level differences along with geographical origin clustering of various yeast isolates which otherwise could not be revealed through conventional characterization. URP markers were found to be best for revealing the genetic polymorphism hidden among forty-three yeast isolates followed by delta typing, RAPD and ISSR. In the above study, URP 6R and URP 9F were found to be species specific thereby producing specific banding pattern for a specific species.

Community science image libraries offer a massive, but largely untapped, source of observational data for phenological research. The iNaturalist platform offers a particularly rich archive, containing more than 49 million verifiable, georeferenced, open access images, encompassing seven continents and over 278,000 species. A critical limitation preventing scientists from taking full advantage of this rich data source is labor. Each image must be manually inspected and categorized by phenophase, which is both time-intensive and costly. Consequently, researchers may only be able to use a subset of the total number of images available in the database. While iNaturalist has the potential to yield enough data for high-resolution and spatially extensive studies, it requires more efficient tools for phenological data extraction. A promising solution is automation of the image annotation process using deep learning. Recent innovations in deep learning have made these open-source tools accessible to a general research audience. However, it is unknown whether deep learning tools can accurately and efficiently annotate phenophases in community science images. Here, we train a convolutional neural network (CNN) to annotate images of Alliaria petiolata into distinct phenophases from iNaturalist and compare the performance of the model with non-expert human annotators. We demonstrate that researchers can successfully employ deep learning techniques to extract phenological information from community science images. A CNN classified two-stage phenology (flowering and non-flowering) with 95.9% accuracy and classified four-stage phenology (vegetative, budding, flowering, and fruiting) with 86.4% accuracy. The overall accuracy of the CNN did not differ from humans (p = 0.383), although performance varied across phenophases. We found that a primary challenge of using deep learning for image annotation was not related to the model itself, but instead in the quality of the community science images. Up to 4% of A. petiolata images in iNaturalist were taken from an improper distance, were physically manipulated, or were digitally altered, which limited both human and machine annotators in accurately classifying phenology. Thus, we provide a list of photography guidelines that could be included in community science platforms to inform community scientists in the best practices for creating images that facilitate phenological analysis.

New advances on the Brettanomyces bruxellensis biofilm mode of life

This paper presents a brief overview of vibration-based damage identification studies based on Deep Learning (DL) in civil engineering structures. The presence, type, size, and propagation of structural damage on civil infrastructure have always been a topic of research. In the last couple of decades, there has been a significant shift in the damage detection paradigm when the advancements in sensing and computing technologies met with the ever-expanding use of artificial neural network algorithms. Machine-Learning (ML) tools enabled researchers to implement more feasible and faster tools in damage detection applications. When an artificial neural network has more than three layers, it is typically considered as a ?deep? learning network. Being an important accomplishment of the ML era, DL tools enable complex systems which are made of several layers to learn implementations of data with outstanding categorization and compartmentalization capability. In fact, with proper training, a DL tool can operate directly with the unprocessed raw data and help the algorithm produce output data. Competitive capabilities like this led DL algorithms perform very well in complicated problems by dividing a relatively large problem into much smaller and more manageable portions. Specifically for damage identification and localization on civil infrastructure, Convolutional Neural Networks (CNNs) and Unsupervised Pretrained Networks (UPNs) are the known DL tools published in the literature. This paper presents an overview of these studies.

Deep learning (also known as deep structured learning, hierarchical learning, or deep machine learning) is a branch of machine learning based on a set of algorithms that attempt to model high-level abstractions in data by using a deep graph with multiple processing layers, composed of multiple linear and nonlinear transformations. Deep learning has been characterized as a class of machine learning algorithms with the following characteristics [257]: • They use a cascade of many layers of nonlinear processing units for feature extraction and transformation. Each successive layer uses the output from the previous layer as input. The algorithms may be supervised or unsupervised and applications include pattern analysis (unsupervised) and classification (supervised). • They are based on the (unsupervised) learning of multiple levels of features or representations of the data. Higher-level features are derived from lower-level features to form a hierarchical representation. • They are part of the broader machine learning field of learning representations of data. • They learn multiple levels of representations that correspond to different levels of abstraction; the levels form a hierarchy of concepts. These definitions have in common: multiple layers of nonlinear processing units and the supervised or unsupervised learning of feature representations in each layer, with the layers forming a hierarchy from low-level to high-level features. Various deep learning architectures such as deep neural networks, convolutional deep neural networks, deep belief networks (DBN), and recurrent neural networks have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition, and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. In this chapter we start in Section 7.1 with an introduction, giving a brief history of this field, the relevant literature, and its applications. Then we study some basic concepts of deep learning such as convolutional neural networks, recurrent neural networks, backpropagation algorithm, restricted Boltzmann machines, and deep learning networks in Section 7.2. Then we illustrate three examples for Apache Spark implementation for mobile big data (MBD), user moving pattern extraction, and combination with nonparametric Bayesian learning, respectively, in Sections 7.3 through 7.5. Finally, we have summary in Section 7.6.

Alzheimer's disease (AD) is a progressive brain disease. The goal of this study is to provide a new computer-vision based technique to detect it in an efficient way. The brain-imaging data of 98AD patients and 98 healthy controls was collected using data augmentation method. Then, convolutional neural network (CNN) was used, CNN is the most successful tool in deep learning. An 8-layer CNN was created with optimal structure obtained by experiences. Three activation functions (AFs): sigmoid, rectified linear unit (ReLU), and leaky ReLU. The three pooling-functions were also tested: average pooling, max pooling, and stochastic pooling. The numerical experiments demonstrated that leaky ReLU and max pooling gave the greatest result in terms of performance. It achieved a sensitivity of 97.96%, a specificity of 97.35%, and an accuracy of 97.65%, respectively. In addition, the proposed approach was compared with eight state-of-the-art approaches. The method increased the classification accuracy by approximately 5% compared to state-of-the-art methods.

Genetic diversity in four populations of the hill stream teleost Barilius bendelisis, a threatened fish species in India, from Ramnagar (Kosi), Srinagar (Alal<nanda), Dehradun (Yamuna) and Bhimtal (Chanfee) Uttarakhand, was studied using Random Amplified Polymorphic DNA (RAPD) markers for the first time in India. Four selective primers provided distinct and consistent RAPD profiles in all the four populations. Upon screening 40 random decamer primers four primers giving reproducible results were selected and used for generating DNA fingerprints from a total of 48 fish samples collected from the above-mentioned streams. The bands in the range 666-4830 bp were scored for consistent results. A total of 30 loci were scored, 12 of which were found to be polymorphic in nature. The mean values of Nei's gene diversity and Shannon's information index were found 0.1899 and 0.2617 respectively indicating an average level of genetic diversity in the studied samples of the threatened species. The RAPD profiles generated by all the four primers revealed varying degrees of polymorphism, ranging from 25% (primer OPH03) to 45.45% (primer OPA09). The overall oroportion of polymorphic bands and genetic diversity (h) among the four populations was observed 40% and 0.1899.

Fusarium rot disease of Tuberose (Polianthes tuberosa L.), caused by the vascular wilt pathogen Fusarium oxysporum, is a major production constraint in Tuberose growing regions of Iran. Samples collected from different fields in Dezful Khuzestan province, and 110 isolates of F. oxysporum recovered. Genetic diversity among F. oxysporum isolates determined using vegetative compatibility groups (VCGs) and Random Amplified Polymorphic DNA (RAPD) techniques. In the first experiments, 45 isolates selected randomly to VCG test. Nit mutants generated on MMC and Czapeck media each containing 3% KClO3. The isolates assigned to into four VCG groups. These isolates also characterized by RAPD technique into six genetic groups at 61% similarity level. Cluster analysis of the RAPD data showed a close agreement with VCG grouping in some cases. In addition, Pathogenicity test revealed that all isolates were pathogenic. This research is the first report of genetic diversity of F. oxysporum on Tuberose in Iran.

This study investigates the effectiveness of Artificial Intelligence (AI) and Deep Learning (DL) tools in enhancing academic journal writing skills among students in the Arabic Language Education program at UIN Maulana Malik Ibrahim Malang. Utilizing a mixed-methods approach within a quasi-experimental design, the research involved 90 final-year students divided into experimental and control groups. The intervention group employed AI-based tools such as ChatGPT, Grammarly, and Quilbolt throughout the writing process, while the control group relied on conventional methods. Data were collected through pre-test and post-test assessments, reflective journals, and structured questionnaires. Quantitative results showed a statistically significant improvement in writing performance among students who used AI tools, with large effect sizes (Cohen’s d &gt; 1.0). Qualitative findings revealed that students engaged critically with AI outputs, valued teacher feedback, and developed ethical awareness regarding authorship and originality. The integration of AI tools also increased student confidence, enhanced writing fluency, and promoted autonomous learning. However, limitations in semantic precision and rhetorical fit especially in theology-specific content necessitated human revision. These findings affirm the role of AI as a cognitive scaffold in academic writing and highlight the need for culturally responsive, ethically guided AI integration in language teacher education.Keywords: Artificial Intelligence, Academic Writing, Deep Learning, Arabic Education, Cognitive Scaffold