A distance-based kernel for classification via Support Vector Machines.

Support Vector Machines (SVMs) are a type of supervised machine learning algorithm widely used for classification tasks. In contrast to traditional methods that split the data into separate training and testing sets, here we propose an innovative approach where subsets of the original data are randomly selected to train the model multiple times. This iterative training process aims to identify a representative data subset, leading to improved inferences about the population. Additionally, we introduce a novel distance-based kernel specifically designed for binary-type features based on a similarity matrix that efficiently handles both binary and multi-class classification problems. Computational experiments on publicly available datasets of varying sizes demonstrate that our proposed method significantly outperforms existing approaches in terms of classification accuracy. Furthermore, the distance-based kernel achieves superior performance compared to other well-known kernels from the literature and those used in previous studies on the same datasets. These findings validate the effectiveness of our proposed classification method and distance-based kernel for SVMs. By leveraging random subset selection and a unique kernel design, we achieve notable improvements in classification accuracy. These results have significant implications for diverse classification problems in Machine Learning and data analysis.

A comparative study of the cortical function during the interpretation of algorithms in pseudocode and the solution of first-order algebraic equations.

This study intends to determine whether similarities of the functioning of the cerebral cortex exist, modeled as a graph, during the execution of mathematical tasks and programming related tasks. The comparison is done using network parameters and during the development of computer programming tasks and the solution of first-order algebraic equations. For that purpose, electroencephalographic recordings (EEG) were made with a volunteer group of 16 students of systems engineering of Universidad del Norte in Colombia, while they were performing computer programming tasks and solving first-order algebraic equations with three levels of difficulty. Then, based on the Synchronization Likelihood method, graph models of functional cortical networks were developed, whose parameters of Small-Worldness (SWN), global(Eg) and local (El) efficiency were compared between both types of tasks. From this study, it can be highlighted, first, the novelty of studying cortical function during the solution of algebraic equations and during programming tasks; second, significant differences between both types of tasks observed only in the delta and theta bands. Likewise, the differences between simpler mathematical tasks with the other levels in both types of tasks; third, the Brodmann areas 21 and 42, associated with auditory sensory processing, can be considered as differentiating elements of programming tasks; as well as Brodmann area 8, during equation solving.

A specific structure and high richness characterize intestinal microbiota of HIV-exposed seronegative individuals.

Intestinal microbiota facilitates food breakdown for energy metabolism and influences the immune response, maintaining mucosal homeostasis. Overall, HIV infection is associated with intestinal dysbiosis and immune activation, which has been related to seroconversion in HIV-exposed individuals. However, it is unclear whether microbiota dysbiosis is the cause or the effect of immune alterations and disease progression or if it could modulate the risk of acquiring the HIV infection. We characterize the intestinal microbiota and determine its association with immune regulation in HIV-exposed seronegative individuals (HESN), HIV-infected progressors (HIV+), and healthy control (HC) subjects. For this, feces and blood were collected. The microbiota composition of HESN showed a significantly higher alpha (p = 0.040) and beta diversity (p = 0.006) compared to HC, but no differences were found compared to HIV+. A lower Treg percentage was observed in HESN (1.77%) than HC (2.98%) and HIV+ (4.02%), with enrichment of the genus Butyrivibrio (p = 0.029) being characteristic of this profile. Moreover, we found that Megasphaera (p = 0.017) and Victivallis (p = 0.0029) also are enriched in the microbiota composition in HESN compared to HC and HIV+ subjects. Interestingly, an increase in Succinivibrio and Prevotella, and a reduction in Bacteroides genus, which is typical of HIV-infected individuals, were observed in both HESN and HIV+, compared to HC. Thus, HESNs have a microbiota profile, similar to that observed in HIV+, most likely because HESN are cohabiting with their HIV+ partners.

A new pipeline for structural characterization and classification of RNA-Seq microbiome data.

BACKGROUND: High-throughput sequencing enables the analysis of the composition of numerous biological systems, such as microbial communities. The identification of dependencies within these systems requires the analysis and assimilation of the underlying interaction patterns between all the variables that make up that system. However, this task poses a challenge when considering the compositional nature of the data coming from DNA-sequencing experiments because traditional interaction metrics (e.g., correlation) produce unreliable results when analyzing relative fractions instead of absolute abundances. The compositionality-associated challenges extend to the classification task, as it usually involves the characterization of the interactions between the principal descriptive variables of the datasets. The classification of new samples/patients into binary categories corresponding to dissimilar biological settings or phenotypes (e.g., control and cases) could help researchers in the development of treatments/drugs. RESULTS: Here, we develop and exemplify a new approach, applicable to compositional data, for the classification of new samples into two groups with different biological settings. We propose a new metric to characterize and quantify the overall correlation structure deviation between these groups and a technique for dimensionality reduction to facilitate graphical representation. We conduct simulation experiments with synthetic data to assess the proposed method's classification accuracy. Moreover, we illustrate the performance of the proposed approach using Operational Taxonomic Unit (OTU) count tables obtained through 16S rRNA gene sequencing data from two microbiota experiments. Also, compare our method's performance with that of two state-of-the-art methods. CONCLUSIONS: Simulation experiments show that our method achieves a classification accuracy equal to or greater than 98% when using synthetic data. Finally, our method outperforms the other classification methods with real datasets from gene sequencing experiments.

A Radiogenomics Ensemble to Predict EGFR and KRAS Mutations in NSCLC.

Lung cancer causes more deaths globally than any other type of cancer. To determine the best treatment, detecting EGFR and KRAS mutations is of interest. However, non-invasive ways to obtain this information are not available. Furthermore, many times there is a lack of big enough relevant public datasets, so the performance of single classifiers is not outstanding. In this paper, an ensemble approach is applied to increase the performance of EGFR and KRAS mutation prediction using a small dataset. A new voting scheme, Selective Class Average Voting (SCAV), is proposed and its performance is assessed both for machine learning models and CNNs. For the EGFR mutation, in the machine learning approach, there was an increase in the sensitivity from 0.66 to 0.75, and an increase in AUC from 0.68 to 0.70. With the deep learning approach, an AUC of 0.846 was obtained, and with SCAV, the accuracy of the model was increased from 0.80 to 0.857. For the KRAS mutation, both in the machine learning models (0.65 to 0.71 AUC) and the deep learning models (0.739 to 0.778 AUC), a significant increase in performance was found. The results obtained in this work show how to effectively learn from small image datasets to predict EGFR and KRAS mutations, and that using ensembles with SCAV increases the performance of machine learning classifiers and CNNs. The results provide confidence that as large datasets become available, tools to augment clinical capabilities can be fielded.

Annual microbial community dynamics in a full-scale anaerobic sludge digester from a wastewater treatment plant in Colombia.

Anaerobic digestion is a microbe-driven process widely applied to treat activated sludge from municipal wastewater treatment plants. It is one of the most efficient solutions for sludge reduction along with biogas production. However, the knowledge of the microbial consortium involved in this process is still unknown in full-scale anaerobic digesters from Latin America. This study aimed to elucidate the dynamics of the microbial community of a full-scale anaerobic digester for a year using 16S rDNA amplicon sequencing with the Illumina Miseq platform. The results showed fluctuations in the frequencies of dominant phyla with a decrease of Proteobacteria and Bacteroidetes after a temporary suspension of anaerobic digester. The core community was affiliated with bacterial phyla Firmicutes, Actinobacteria, Proteobacteria, and Chloroflexi. The core community was represented by 154 OTUs that accounted for 74% of all the processed reads. The Anaerolineaceae family, within Chloroflexi phylum, was the most frequently observed taxonomic group in all samples analyzed. Despite the microbial fluctuations, the biogas production was stable over the studied year (average 66% methane production), which might indicate a functional redundancy in the microbial consortium.

Non-Local SVD Denoising of MRI Based on Sparse Representations.

Magnetic Resonance (MR) Imaging is a diagnostic technique that produces noisy images, which must be filtered before processing to prevent diagnostic errors. However, filtering the noise while keeping fine details is a difficult task. This paper presents a method, based on sparse representations and singular value decomposition (SVD), for non-locally denoising MR images. The proposed method prevents blurring, artifacts, and residual noise. Our method is composed of three stages. The first stage divides the image into sub-volumes, to obtain its sparse representation, by using the KSVD algorithm. Then, the global influence of the dictionary atoms is computed to upgrade the dictionary and obtain a better reconstruction of the sub-volumes. In the second stage, based on the sparse representation, the noise-free sub-volume is estimated using a non-local approach and SVD. The noise-free voxel is reconstructed by aggregating the overlapped voxels according to the rarity of the sub-volumes it belongs, which is computed from the global influence of the atoms. The third stage repeats the process using a different sub-volume size for producing a new filtered image, which is averaged with the previously filtered images. The results provided show that our method outperforms several state-of-the-art methods in both simulated and real data.

Study of the impact of long-duration space missions at the International Space Station on the astronaut microbiome.

Over the course of a mission to the International Space Station (ISS) crew members are exposed to a number of stressors that can potentially alter the composition of their microbiomes and may have a negative impact on astronauts' health. Here we investigated the impact of long-term space exploration on the microbiome of nine astronauts that spent six to twelve months in the ISS. We present evidence showing that the microbial communities of the gastrointestinal tract, skin, nose and tongue change during the space mission. The composition of the intestinal microbiota became more similar across astronauts in space, mostly due to a drop in the abundance of a few bacterial taxa, some of which were also correlated with changes in the cytokine profile of crewmembers. Alterations in the skin microbiome that might contribute to the high frequency of skin rashes/hypersensitivity episodes experienced by astronauts in space were also observed. The results from this study demonstrate that the composition of the astronauts' microbiome is altered during space travel. The impact of those changes on crew health warrants further investigation before humans embark on long-duration voyages into outer space.

Split and Merge Watershed: a two-step method for cell segmentation in fluorescence microscopy images.

The development of advanced techniques in medical imaging has allowed scanning of the human body to microscopic levels, making research on cell behavior more complex and more in-depth. Recent studies have focused on cellular heterogeneity since cell-to-cell differences are always present in the cell population and this variability contains valuable information. However, identifying each cell is not an easy task because, in the images acquired from the microscope, there are clusters of cells that are touching one another. Therefore, the segmentation stage is a problem of considerable difficulty in cell image processing. Although several methods for cell segmentation are described in the literature, they have drawbacks in terms of over-segmentation, under-segmentation or misidentification. Consequently, our main motivation in studying cell segmentation was to develop a new method to achieve a good tradeoff between accurately identifying all relevant elements and not inserting segmentation artifacts. This article presents a new method for cell segmentation in fluorescence microscopy images. The proposed approach combines the well-known Marker-Controlled Watershed algorithm (MC-Watershed) with a new, two-step method based on Watershed, Split and Merge Watershed (SM-Watershed): in the first step, or split phase, the algorithm identifies the clusters using inherent characteristics of the cell, such as size and convexity, and separates them using watershed. In the second step, or the merge stage, it identifies the over-segmented regions using proper features of the cells and eliminates the divisions. Before applying our two-step method, the input image is first preprocessed, and the MC-Watershed algorithm is used to generate an initial segmented image. However, this initial result may not be suitable for subsequent tasks, such as cell count or feature extraction, because not all cells are separated, and some cells may be mistakenly confused with the background. Thus, our proposal corrects this issue with its two-step process, reaching a high performance, a suitable tradeoff between over-segmentation and under-segmentation and preserving the shape of the cell, without the need of any labeled data or relying on machine learning processes. The latter is advantageous over state-of-the-art techniques that in order to achieve similar results require labeled data, which may not be available for all of the domains. Two cell datasets were used to validate this approach, and the results were compared with other methods in the literature, using traditional metrics and quality visual assessment. We obtained 90% of average visual accuracy and an F-index higher than 80%. This proposal outperforms other techniques for cell separation, achieving an acceptable balance between over-segmentation and under-segmentation, which makes it suitable for several applications in cell identification, such as virus infection analysis, high-content cell screening, drug discovery, and morphometry.

Gut microbiota components are associated with fixed airway obstruction in asthmatic patients living in the tropics.

Microbiome composition has been associated to several inflammatory diseases, including asthma. There are few studies exploring the relationships of gut microbiota with airway obstruction pheonotypes in adult asthma, especially those living in the tropics. We sought to evaluate the relationships of gut microbiota with the airway obstruction and other variables of interest in asthmatic patients living in the tropics according to three phenotypes: No Airway Obstruction (NAO), Reversible Airway Obstruction (RAO) or Fixed Airway Obstruction (FAO). We found that Streptococcaceae:Streptococcus and Enterobacteriaceae:Escherichia-Shigella consistently discriminated asthmatic individuals suffering FAO from NAO or RAO, plus Veillonellaceae:Megasphaera when comparing FAO and RAO (p < 0.05; FDR < 0.05). In the FAO, the network showing the genus relations was less complex and interconnected. Several Rumminococcaceae, Lachnospiraceae and Clostridiales were enriched in patients with low specific IgE levels to mites and Ascaris. All patients shared a common exposure framework; control medication usage and smoking habit were uncommon and equally distributed between them. In conclusion, in this tropical asthmatic population, components of human gut microbiota are associated with the presence of a FAO phenotype and lower specific IgE response to mites and Ascaris.

Effect of temperature in bursting of thalamic reticular cells / Efecto de la temperatura en las descargas en ráfaga de la célula del núcleo reticular del tálamo

Abstract Objective: To show the relation between the four parameters associated to bursting discharges of the thalamic reticular cells (TRNn): the maximum firing frequency (fmax) and the temperature at which it occurs (Tfmax), the range of temperatures defined as the full width at half maximum (∆Th) and the maximum specific low threshold calcium conductance (GT). Materials and Methods: In order to simulate the TRNn bursting activity, a computational simulation model was implemented using the NEURON software, which incorporates morphological and electrophysiological data, and stimuli properties closely related to reality. Results: It was found that there are nonlinear relations between the parameters. The fmax frequency follows a quadratic growth with temperature and tends asymptotically towards a limit value with the maximum calcium conductance. In the same manner, ∆Th increases until reaching a limit value as function of fmax and GT. However, the increment per frequency unit is bigger than the increment per conductance unit. Conclusions: Four equations were obtained that model the relations between the parameters associated to bursting discharges of the TRNn in rats and other neurons with similar characteristics in different animal species.

Resumen Objetivo: Mostrar la relación entre los cuatro parámetros asociados a las descargas en ráfaga de las neuronas del núcleo reticular del tálamo (TRNn): la frecuencia máxima de descarga (fmax) y la temperatura a la cual se produce (Tfmax), el rango de temperaturas definido como ancho a media altura (∆Th) y la conductancia máxima de calcio de bajo umbral (GT). Materiales y métodos: Para simular las descargas en ráfaga de las TRNn se implementó un modelo de simulación computacional usando el software NEURON, que incorpora datos morfológicos, electrofisiologicos y las propiedades de los estímulos en estrecha relación con la realidad. Resultados: Se encontraron relaciones no lineales entre los parámetros. La frecuencia fmax crece de forma cuadrática con la temperatura y tiende asintóticamente a un valor límite con la conductancia. Así mismo, ∆Th también se incrementan hasta alcanzar un valor límite en función de fmax y GT. No obstante, es mayor el incremento por cada unidad de frecuencia que por cada unidad de conductancia. Conclusiones: Se obtuvieron cuatro ecuaciones que modelan las relaciones entre los pará- metros asociados a las descargas en ráfaga de las neuronas TRN en ratas y otras neuronas con características similares en diferentes especies animales.

A Lachnospiraceae-dominated bacterial signature in the fecal microbiota of HIV-infected individuals from Colombia, South America.

HIV infection has a tremendous impact on the immune system's proper functioning. The mucosa-associated lymphoid tissue (MALT) is significantly disarrayed during HIV infection. Compositional changes in the gut microbiota might contribute to the mucosal barrier disruption, and consequently to microbial translocation. We performed an observational, cross-sectional study aimed at evaluating changes in the fecal microbiota of HIV-infected individuals from Colombia. We analyzed the fecal microbiota of 37 individuals via 16S rRNA gene sequencing; 25 HIV-infected patients and 12 control (non-infected) individuals, which were similar in body mass index, age, gender balance and socioeconomic status. To the best of our knowledge, no such studies have been conducted in Latin American countries. Given its compositional nature, microbiota data were normalized and transformed using Aitchison's Centered Log-Ratio. Overall, a change in the network structure in HIV-infected patients was revealed by using the SPIEC-EASI MB tool. Genera such as Blautia, Dorea, Yersinia, Escherichia-Shigella complex, Staphylococcus, and Bacteroides were highly relevant in HIV-infected individuals. Differential abundance analysis by both sparse Partial Least Square-Discriminant Analysis and Random Forest identified a greater abundance of Lachnospiraceae-OTU69, Blautia, Dorea, Roseburia, and Erysipelotrichaceae in HIV-infected individuals. We show here, for the first time, a predominantly Lachnospiraceae-based signature in HIV-infected individuals.

Digital Image Analysis of Cells and Computational Tools for the Study of Mechanism of RSV Entry to Human Bronchial Epithelium.

this paper presents a research proposal which has been developed as a doctoral thesis in the PhD program in Computer Systems Engineering at the Universidad del Norte since August 2015. This research focuses on the analysis of cell images of the human bronchial epithelium infected with the Respiratory Syncytial Virus in order to understand the mechanisms of entry of the virus into the human body. Due to the large amount of information that is processed, it is necessary to use computational tools to finally differentiate between infected and uninfected cells.

Thalamic reticular cells firing modes and its dependency on the frequency and amplitude ranges of the current stimulus.

The neurons of the Thalamic Reticular Nucleus (TRNn) respond to inputs in two activity modes called burst and tonic firing and both can be observed in different physiological states. The functional states of the thalamus depend in part on the properties of synaptic transmission between the TRNn and the thalamocortical and corticothalamic neurons. A dendrite can receive inhibitory and excitatory postsynaptic potentials. The novelties presented in this paper can be summarized as follows: First, it shows, through a computational simulation, that the burst and tonic firings observed in the TRNn soma could be explained as a product of random synaptic inputs on the distal dendrites, the tonic firings are generated by random excitatory stimuli, and the burst firings are generated by two different types of stimuli: inhibitory random stimuli, and a combination of inhibitory (from TRNn) and excitatory (from corticothalamic and thalamocortical neurons) random stimuli; second, according to in vivo recordings, we have found that the burst observed in the TRNn soma has graduate properties that are proportional to the stimuli frequency; and third, a novel method for showing in a quantitative manner the accelerando-decelerando pattern is proposed.

Teaching and learning the Hodgkin-Huxley model based on software developed in NEURON's programming language hoc.

BACKGROUND: We present a software tool called SENB, which allows the geometric and biophysical neuronal properties in a simple computational model of a Hodgkin-Huxley (HH) axon to be changed. The aim of this work is to develop a didactic and easy-to-use computational tool in the NEURON simulation environment, which allows graphical visualization of both the passive and active conduction parameters and the geometric characteristics of a cylindrical axon with HH properties. RESULTS: The SENB software offers several advantages for teaching and learning electrophysiology. First, SENB offers ease and flexibility in determining the number of stimuli. Second, SENB allows immediate and simultaneous visualization, in the same window and time frame, of the evolution of the electrophysiological variables. Third, SENB calculates parameters such as time and space constants, stimuli frequency, cellular area and volume, sodium and potassium equilibrium potentials, and propagation velocity of the action potentials. Furthermore, it allows the user to see all this information immediately in the main window. Finally, with just one click SENB can save an image of the main window as evidence. CONCLUSIONS: The SENB software is didactic and versatile, and can be used to improve and facilitate the teaching and learning of the underlying mechanisms in the electrical activity of an axon using the biophysical properties of the squid giant axon.