MSDeepAMR: antimicrobial resistance prediction based on deep neural networks and transfer learning.

Introduction: Antimicrobial resistance (AMR) is a global health problem that requires early and effective treatments to prevent the indiscriminate use of antimicrobial drugs and the outcome of infections. Mass Spectrometry (MS), and more particularly MALDI-TOF, have been widely adopted by routine clinical microbiology laboratories to identify bacterial species and detect AMR. The analysis of AMR with deep learning is still recent, and most models depend on filters and preprocessing techniques manually applied on spectra. Methods: This study propose a deep neural network, MSDeepAMR, to learn from raw mass spectra to predict AMR. MSDeepAMR model was implemented for Escherichia coli, Klebsiella pneumoniae, and Staphylococcus aureus under different antibiotic resistance profiles. Additionally, a transfer learning test was performed to study the benefits of adapting the previously trained models to external data. Results: MSDeepAMR models showed a good classification performance to detect antibiotic resistance. The AUROC of the model was above 0.83 in most cases studied, improving the results of previous investigations by over 10%. The adapted models improved the AUROC by up to 20% when compared to a model trained only with external data. Discussion: This study demonstrate the potential of the MSDeepAMR model to predict antibiotic resistance and their use on external MS data. This allow the extrapolation of the MSDeepAMR model to de used in different laboratories that need to study AMR and do not have the capacity for an extensive sample collection.

Prospective Cohort Study Identifies Medical Predictors of Treatment-Related Oral Toxicities in Oral and Oropharyngeal Cancer Patients.

The dental treatment of patients with oral cavity and oropharyngeal squamous cell carcinoma (OOPSCC) may be challenging for dentists. This study aimed to characterize systemic changes in patients with OOPSCC undergoing dental treatment prior to cancer therapy, with a specific focus on laboratory assessments. The primary objectives included identifying potential adverse events, such as infections or bleeding, resulting from dental procedures. Additionally, the study aimed to correlate baseline patient characteristics with treatment-related toxicities. This was a prospective cohort study that included 110 OOPSCC patients referred to the Dental Oncology Service at São Paulo State Cancer Institute, Brazil, between November/2019 and December/2020. Comorbidities, sociodemographic data, medication in use, cancer treatment-related toxicities, and altered laboratory tests results were correlated. The most common comorbidities and altered laboratory results were hypertension, dyslipidemia, diabetes, as well as elevated levels of C-reactive protein, hemoglobin, and hematocrit. Toxicities exhibited a progressive pattern over time, encompassing oral mucositis (OM), xerostomia, dysphagia, dysgeusia, trismus, and radiodermatitis. No correlation between comorbidities and cancer treatment-related toxicities, a positive correlation between medications in use and OM, and a negative correlation between medications and dysgeusia were found. OM was associated with altered thyroxine (T4) and free thyroxine (FT4), calcium, urea, creatinine, alkaline phosphatase, and syphilis. Family income and housing were OM predictors. Altered T4/FT4/urea/calcium/alkaline phosphatase/creatinine/syphilis may be useful clinical predictors of OM. Despite the elevated prevalence of comorbidities and abnormal laboratory findings, dental treatment prior to cancer treatment yielded no adverse events.

Evaluating deep learning predictions for COVID-19 from X-ray images using leave-one-out predictive densities.

Early detection of the COVID-19 virus is an important task for controlling the spread of the pandemic. Imaging techniques such as chest X-ray are relatively inexpensive and accessible, but its interpretation requires expert knowledge to evaluate the disease severity. Several approaches for automatic COVID-19 detection using deep learning techniques have been proposed. While most approaches show high accuracy on the COVID-19 detection task, there is not enough evidence on external evaluation for this technique. Furthermore, data scarcity and sampling biases make difficult to properly evaluate model predictions. In this paper, we propose stochastic gradient Langevin dynamics (SGLD) to take into account the model uncertainty. Four different deep learning architectures are trained using SGLD and compared to their baselines using stochastic gradient descent. The model uncertainties are also evaluated according to their convergence properties and the leave-one-out predictive densities. The proposed approach is able to reduce overconfidence of the baseline estimators while also retaining predictive accuracy for the best-performing cases.

Novel metabolomic profile of subjects with non-classic apparent mineralocorticoid excess.

Nonclassic apparent mineralocorticoid excess (NC-AME) is proposed as a novel clinical condition with a mild phenotypic spectrum that ranges from normotension to severe hypertension. This condition is mainly characterized by a high serum cortisol to cortisone ratio (F/E) and concomitant low cortisone (E), however further metabolic changes in NC-AME have not been studied. A cross-sectional study was performed in a primary-care cohort of 396 Chilean subjects, which were classified in two groups: NC-AME (n = 28) and healthy controls (n = 27). A discovery study based in untargeted metabolomics assay in serum samples from both groups was performed by UPLC-Q-TOF/MS. Global metabolomic variations were assayed by principal component analysis and further compared by orthogonal partial least-squares discriminant analysis (OPLS-DA). NC-AME subjects exhibited higher values of blood pressure, fractional excretion of potassium, and lower plasma renin activity and urinary sodium to potassium ratio. Metabolomic analyses showed 36 differentially regulated metabolites between NC-AME and control subjects. A ROC curve analyses identified eight metabolites with high discriminatory capacity between NC-AME and control subjects. Moreover, gamma-L-glutamyl-L-methionine sulfoxide and 5-sulfoxymethylfurfural, exhibited significant association with cortisone, which are potential biomarkers of NC-AME, however further assays should elucidate its biological role in setup and progression of this phenotype.

Effects of Interleukin-1ß in Glycinergic Transmission at the Central Amygdala.

Interleukin-1ß (IL-1ß) is an important cytokine that modulates peripheral and central pain sensitization at the spinal level. Among its effects, it increases spinal cord excitability by reducing inhibitory Glycinergic and GABAergic neurotransmission. In the brain, IL-1ß is released by glial cells in regions associated with pain processing during neuropathic pain. It also has important roles in neuroinflammation and in regulating NMDA receptor activity required for learning and memory. The modulation of glycine-mediated inhibitory activity via IL-1ß may play a critical role in the perception of different levels of pain. The central nucleus of the amygdala (CeA) participates in receiving and processing pain information. Interestingly, this nucleus is enriched in the regulatory auxiliary glycine receptor (GlyR) ß subunit (ßGlyR); however, no studies have evaluated the effect of IL-1ß on glycinergic neurotransmission in the brain. Hence, we hypothesized that IL-1ß may modulate GlyR-mediated inhibitory activity via interactions with the ßGlyR subunit. Our results show that the application of IL-1ß (10 ng/ml) to CeA brain slices has a biphasic effect; transiently increases and then reduces sIPSC amplitude of CeA glycinergic currents. Additionally, we performed molecular docking, site-directed mutagenesis, and whole-cell voltage-clamp electrophysiological experiments in HEK cells transfected with GlyRs containing different GlyR subunits. These data indicate that IL-1ß modulates GlyR activity by establishing hydrogen bonds with at least one key amino acid residue located in the back of the loop C at the ECD domain of the ßGlyR subunit. The present results suggest that IL-1ß in the CeA controls glycinergic neurotransmission, possibly via interactions with the ßGlyR subunit. This effect could be relevant for understanding how IL-1ß released by glia modulates central processing of pain, learning and memory, and is involved in neuroinflammation.

A Deep Learning Approach to Population Structure Inference in Inbred Lines of Maize.

Analysis of population genetic variation and structure is a common practice for genome-wide studies, including association mapping, ecology, and evolution studies in several crop species. In this study, machine learning (ML) clustering methods, K-means (KM), and hierarchical clustering (HC), in combination with non-linear and linear dimensionality reduction techniques, deep autoencoder (DeepAE) and principal component analysis (PCA), were used to infer population structure and individual assignment of maize inbred lines, i.e., dent field corn (n = 97) and popcorn (n = 86). The results revealed that the HC method in combination with DeepAE-based data preprocessing (DeepAE-HC) was the most effective method to assign individuals to clusters (with 96% of correct individual assignments), whereas DeepAE-KM, PCA-HC, and PCA-KM were assigned correctly 92, 89, and 81% of the lines, respectively. These findings were consistent with both Silhouette Coefficient (SC) and Davies-Bouldin validation indexes. Notably, DeepAE-HC also had better accuracy than the Bayesian clustering method implemented in InStruct. The results of this study showed that deep learning (DL)-based dimensional reduction combined with ML clustering methods is a useful tool to determine genetically differentiated groups and to assign individuals into subpopulations in genome-wide studies without having to consider previous genetic assumptions.

New polymer for removal of wine phenolics: Poly(N-(3-(N-isobutyrylisobutyramido)-3-oxopropyl)acrylamide) (P-NIOA).

The phenolic compounds of wine contribute to color and astringency, also are responsible for the oxidation state and bitterness. Due the importance of these molecules, different techniques have been used to modulate their concentration such as natural or synthetic polymeric agents. Among the polymeric agents, PVPP is one of the most used, but lacks of selectivity and has a limited pH range. Therefore, the aim of this study was the synthesis of a new polymer, poly(N-(3-(N-isobutyrylisobutyramido)-3-oxopropyl)acrylamide) (P-NIOA), for removal of phenolic compounds, as a potential agent for the fining of wine. The new polymer affinity was studied using HPLC-DAD for different polyphenols using PVPP as a control. The results showed that the new polymer has a similar removal as PVPP, but with lower affinity to resveratrol. The interactions established between polymers and polyphenols were studied using computational chemistry methods demonstrating a direct correlation with the experimental affinity data.

Experimental and theoretical binding affinity between polyvinylpolypyrrolidone and selected phenolic compounds from food matrices.

Polyvinylpolypyrrolidone (PVPP) is a fining agent, widely used in winemaking and brewing, whose mode of action in removing phenolic compounds has not been fully characterised. The aim of this study was to evaluate the experimental and theoretical binding affinity of PVPP towards six phenolic compounds representing different types of phenolic species. The interaction between PVPP and phenolics was evaluated in model solutions, where hydroxyl groups, hydrophobic bonding and steric hindrance were characterised. The results of the study indicated that PVPP exhibits high affinity for quercetin and catechin, moderate affinity for epicatechin, gallic acid and lower affinity for 4-methylcatechol and caffeic acid. The affinity has a direct correlation with the hydroxylation degree of each compound. The results show that the affinity of PVPP towards phenols is related with frontier orbitals. This work demonstrates a direct correlation between the experimental affinity and the interaction energy calculations obtained through computational chemistry methods.

pH-dependent nano-capturing of tartaric acid using dendrimers.

The ability of dendrimers to bind to various target molecules through non-covalent interactions was used to capture water soluble organic reagents, such as tartaric acid (TA), from different matrices, i.e. aqueous solutions and wine samples. The influence of the pH, dendrimer type, generation and feeding concentration on the host-guest complexation of TA was investigated. The maximum binding capacity of TA in aqueous solutions was achieved by amine end-capped dendrimers at pH 5. At extreme pH values of 2 and 11, the binding of TA dropped strikingly, demonstrating the pH-dependency underlying the host-guest interactions. The linear correlation between the maximum binding capacity of TA at pH 5 and the number of primary amine groups on the surface of PAMAM and PPI dendrimers strongly indicated that host-guest complex formation between TA and dendrimers is largely dependent on electrostatic interactions. Molecular simulations confirmed the predominant electrostatic nature of the interactions between TA and the amine end-capped dendrimers and also provided important information on the spatial distribution of TA within the PAMAM G5 dendrimer. All these results designate dendrimers as potential nano-capturing systems for the removal/recovery of TA from complex matrices such as wine, industrial waste or fruit juices.

In silico analysis of putative paralytic shellfish poisoning toxins export proteins in cyanobacteria.

Paralytic shellfish poisoning toxins (PSTs) are a family of more than 30 natural alkaloids synthesized by dinoflagellates and cyanobacteria whose toxicity in animals is mediated by voltage-gated Na(+) channel blocking. The export of PST analogues may be through SxtF and SxtM, two putative MATE (multidrug and toxic compound extrusion) family transporters encoded in PSTs biosynthetic gene cluster (sxt). sxtM is present in every sxt cluster analyzed; however, sxtF is only present in the Cylindrospermopsis-Raphidiopsis clade. These transporters are energetically coupled with an electrochemical gradient of proton (H(+)) or sodium (Na(+)) ions across membranes. Because the functional role of PSTs remains unknown and methods for genetic manipulation in PST-producing organisms have not yet been developed, protein structure analyses will allow us to understand their function. By analyzing the sxt cluster of eight PST-producing cyanobacteria, we found no correlation between the presence of sxtF or sxtM and a specific PSTs profile. Phylogenetic analyses of SxtF/M showed a high conservation of SxtF in the Cylindrospermopsis-Raphidiopsis clade, suggesting conserved substrate affinity. Two domains involved in Na(+) and drug recognition from NorM proteins (MATE family) of Vibrio parahaemolyticus and V. cholerae are present in SxtF/M. The Na(+) recognition domain was conserved in both SxtF/M, indicating that Na(+) can maintain the role as a cation anti-transporter. Consensus motifs for toxin binding differed between SxtF and SxtM implying differential substrate binding. Through protein modeling and docking analysis, we found that there is no marked affinity between the recognition domain and a specific PST analogue. This agrees with our previous results of PST export in R. brookii D9, where we observed that the response to Na(+) incubation was similar to different analogues. These results reassert the hypothesis regarding the involvement of Na(+) in toxin export, as well as the motifs L(398)XGLQD(403) (SxtM) and L(390)VGLRD(395) (SxtF) in toxin recognition.

Biosynthesis of methoxypyrazines: elucidating the structural/functional relationship of two Vitis viniferaO-methyltransferases capable of catalyzing the putative final step of the biosynthesis of 3-alkyl-2-methoxypyrazine.

3-Alkyl-2-methoxypyrazines (MPs) are an important food constituent and have been associated with detrimental herbaceous flavors in red wines by consumers and the wine industry. The Vitis vinifera genes O-methyltransferase 1 and 2 (VvOMT1 and VvOMT2) have been isolated in the grapevine cultivar Carmenere. These genes encode S-adenosyl-l-methionine (SAM)-dependent O-methyltransferases, which have the ability to methylate 3-alkyl-2-hydroxypyrazines (HPs)-the putative final step in MPs production. Atomic studies were performed in order to explain the differences in these VvOMT activities through their structural/functional relationship in MPs biosynthesis. Differences in enthalpy energy observed between the proteins may be due to changes of equivalent residues in the active sites of VvOMT1 (F319, L322) and VvOMT2 (L319, V322). However, docking simulations and QM/MM analyses described how residues H272 and M182 could explain the main functional differentiation observed between VvOMT1 and VvOMT2 through steric impediment, which limits the formation of the transition state in enzymes encoded by VvOMT2. Therefore, this finding could explain the decreasing catalytic efficiency observed for VvOMT2.

Software antropmeter, una nueva herramienta para análisis facial / Antropmeter software, a new tool for facial analysis

El análisis de las dimensiones y proporciones faciales es necesario en distintos ámbitos de la odontoestomatología y de la antropología física. En este informe presentamos el software Antropmeter, diseñado para realizar análisis de dimensiones y proporciones faciales, en base a fotografías estandarizadas, de fácil manejo por parte del clínico y de utilidad en análisis faciales estéticos y antropológicos.

The dimensions and facial proportions analysis are necessary in different areas of the odontostomatology and physical anthropology practice. In this report we present the Antropmeter software, designed to carry out analysis of dimensions and facial proportions, based on standardized pictures, of easy handling on the part of the clinical one and of utility in aesthetic and anthropological facial analysis.