Physical Activity Detection for Diabetes Mellitus Patients Using Recurrent Neural Networks.

Diabetes mellitus (DM) is a persistent metabolic disorder associated with the hormone insulin. The two main types of DM are type 1 (T1DM) and type 2 (T2DM). Physical activity plays a crucial role in the therapy of diabetes, benefiting both types of patients. The detection, recognition, and subsequent classification of physical activity based on type and intensity are integral components of DM treatment. The continuous glucose monitoring system (CGMS) signal provides the blood glucose (BG) level, and the combination of CGMS and heart rate (HR) signals are potential targets for detecting relevant physical activity from the BG variation point of view. The main objective of the present research is the developing of an artificial intelligence (AI) algorithm capable of detecting physical activity using these signals. Using multiple recurrent models, the best-achieved performance of the different classifiers is a 0.99 area under the receiver operating characteristic curve. The application of recurrent neural networks (RNNs) is shown to be a powerful and efficient solution for accurate detection and analysis of physical activity in patients with DM. This approach has great potential to improve our understanding of individual activity patterns, thus contributing to a more personalized and effective management of DM.

Selenosugars targeting the infective stage of Trypanosoma brucei with high selectivity.

Earlier evidences showed that diglycosyl diselenides are active against the infective stage of African trypanosomes (top hits IC50 0.5 and 1.5 µM) but poorly selective (selectivity index <10). Here we extended the study to 33 new seleno-glycoconjugates with the aim to improve potency and selectivity. Three selenoglycosides and three glycosyl selenenylsulfides displayed IC50 against bloodstream Trypanosoma brucei in the sub-µM range (IC50 0.35-0.77 µM) and four of them showed an improved selectivity (selectivity index >38-folds vs. murine and human macrohages). For the glycosyl selenylsulfides, the anti-trypanosomal activity was not significantly influenced by the nature of the moiety attached to the sulfur atom. Except for a quinoline-, and to a minor extent a nitro-derivative, the most selective hits induced a rapid (within 60 min) and marked perturbation of the LMWT-redox homeostasis. The formation of selenenylsulfide glycoconjugates with free thiols has been identified as a potential mechanism involved in this process.

Saturation Transfer Difference NMR and Molecular Docking Interaction Study of Aralkyl-Thiodigalactosides as Potential Inhibitors of the Human-Galectin-3 Protein.

Human Galectin-3 (hGal-3) is a protein that selectively binds to ß-galactosides and holds diverse roles in both normal and pathological circumstances. Therefore, targeting hGal-3 has become a vibrant area of research in the pharmaceutical chemistry. As a step towards the development of novel hGal-3 inhibitors, we synthesized and investigated derivatives of thiodigalactoside (TDG) modified with different aromatic substituents. Specifically, we describe a high-yielding synthetic route of thiodigalactoside (TDG); an optimized procedure for the synthesis of the novel 3,3'-di-O-(quinoline-2-yl)methyl)-TDG and three other known, symmetric 3,3'-di-O-TDG derivatives ((naphthalene-2yl)methyl, benzyl, (7-methoxy-2H-1-benzopyran-2-on-4-yl)methyl). In the present study, using competition Saturation Transfer Difference (STD) NMR spectroscopy, we determined the dissociation constant (Kd) of the former three TDG derivatives produced to characterize the strength of the interaction with the target protein (hGal-3). Based on the Kd values determined, the (naphthalen-2-yl)methyl, the (quinolin-2-yl)methyl and the benzyl derivatives bind to hGal-3 94, 30 and 24 times more strongly than TDG. Then, we studied the binding modes of the derivatives in silico by molecular docking calculations. Docking poses similar to the canonical binding modes of well-known hGal-3 inhibitors have been found. However, additional binding forces, cation-π interactions between the arginine residues in the binding pocket of the protein and the aromatic groups of the ligands, have been established as significant features. Our results offer a molecular-level understanding of the varying affinities observed among the synthesized thiodigalactoside derivatives, which can be a key aspect in the future development of more effective ligands of hGal-3.

Machine Learning on Prediction of Relative Physical Activity Intensity Using Medical Radar Sensor and 3D Accelerometer.

BACKGROUND: One of the most critical topics in sports safety today is the reduction in injury risks through controlled fatigue using non-invasive athlete monitoring. Due to the risk of injuries, it is prohibited to use accelerometer-based smart trackers, activity measurement bracelets, and smart watches for recording health parameters during performance sports activities. This study analyzes the synergy feasibility of medical radar sensors and tri-axial acceleration sensor data to predict physical activity key performance indexes in performance sports by using machine learning (ML). The novelty of this method is that it uses a 24 GHz Doppler radar sensor to detect vital signs such as the heartbeat and breathing without touching the person and to predict the intensity of physical activity, combined with the acceleration data from 3D accelerometers. METHODS: This study is based on the data collected from professional athletes and freely available datasets created for research purposes. A combination of sensor data management was used: a medical radar sensor with no-contact remote sensing to measure the heart rate (HR) and 3D acceleration to measure the velocity of the activity. Various advanced ML methods and models were employed on the top of sensors to analyze the vital parameters and predict the health activity key performance indexes. three-axial acceleration, heart rate data, age, as well as activity level variances. RESULTS: The ML models recognized the physical activity intensity and estimated the energy expenditure on a realistic level. Leave-one-out (LOO) cross-validation (CV), as well as out-of-sample testing (OST) methods, have been used to evaluate the level of accuracy in activity intensity prediction. The energy expenditure prediction with three-axial accelerometer sensors by using linear regression provided 97-99% accuracy on selected sports (cycling, running, and soccer). The ML-based RPE results using medical radar sensors on a time-series heart rate (HR) dataset varied between 90 and 96% accuracy. The expected level of accuracy was examined with different models. The average accuracy for all the models (RPE and METs) and setups was higher than 90%. CONCLUSIONS: The ML models that classify the rating of the perceived exertion and the metabolic equivalent of tasks perform consistently.

AI-Assisted Fatigue and Stamina Control for Performance Sports on IMU-Generated Multivariate Times Series Datasets.

BACKGROUND: Optimal sports performance requires a balance between intensive training and adequate rest. IMUs provide objective, quantifiable data to analyze performance dynamics, despite the challenges in quantifying athlete training loads. The ability of AI to analyze complex datasets brings innovation to the monitoring and optimization of athlete training cycles. Traditional techniques rely on subjective assessments to prevent overtraining, which can lead to injury and underperformance. IMUs provide objective, quantitative data on athletes' physical status during action. AI and machine learning can turn these data into useful insights, enabling data-driven athlete performance management. With IMU-generated multivariate time series data, this paper uses AI to construct a robust model for predicting fatigue and stamina. MATERIALS AND METHODS: IMUs linked to 19 athletes recorded triaxial acceleration, angular velocity, and magnetic orientation throughout repeated sessions. Standardized training included steady-pace runs and fatigue-inducing techniques. The raw time series data were used to train a supervised ML model based on frequency and time-domain characteristics. The performances of Random Forest, Gradient Boosting Machines, and LSTM networks were compared. A feedback loop adjusted the model in real time based on prediction error and bias estimation. RESULTS: The AI model demonstrated high predictive accuracy for fatigue, showing significant correlations between predicted fatigue levels and observed declines in performance. Stamina predictions enabled individualized training adjustments that were in sync with athletes' physiological thresholds. Bias correction mechanisms proved effective in minimizing systematic prediction errors. Moreover, real-time adaptations of the model led to enhanced training periodization strategies, reducing the risk of overtraining and improving overall athletic performance. CONCLUSIONS: In sports performance analytics, the AI-assisted model using IMU multivariate time series data is effective. Training can be tailored and constantly altered because the model accurately predicts fatigue and stamina. AI models can effectively forecast the beginning of weariness before any physical symptoms appear. This allows for timely interventions to prevent overtraining and potential accidents. The model shows an exceptional ability to customize training programs according to the physiological reactions of each athlete and enhance the overall training effectiveness. In addition, the study demonstrated the model's efficacy in real-time monitoring performance, improving the decision-making abilities of both coaches and athletes. The approach enables ongoing and thorough data analysis, supporting strategic planning for training and competition, resulting in optimized performance outcomes. These findings highlight the revolutionary capability of AI in sports science, offering a future where data-driven methods greatly enhance athlete training and performance management.

Detection of Physical Activity Using Machine Learning Methods Based on Continuous Blood Glucose Monitoring and Heart Rate Signals.

Non-coordinated physical activity may lead to hypoglycemia, which is a dangerous condition for diabetic people. Decision support systems related to type 1 diabetes mellitus (T1DM) still lack the capability of automated therapy modification by recognizing and categorizing the physical activity. Further, this desired adaptive therapy should be achieved without increasing the administrative load, which is already high for the diabetic community. These requirements can be satisfied by using artificial intelligence-based solutions, signals collected by wearable devices, and relying on the already available data sources, such as continuous glucose monitoring systems. In this work, we focus on the detection of physical activity by using a continuous glucose monitoring system and a wearable sensor providing the heart rate-the latter is accessible even in the cheapest wearables. Our results show that the detection of physical activity is possible based on these data sources, even if only low-complexity artificial intelligence models are deployed. In general, our models achieved approximately 90% accuracy in the detection of physical activity.

Establishing a Learning Model for Correct Hand Hygiene Technique in a NICU.

The ability of healthcare workers to learn proper hand hygiene has been an understudied area of research. Generally, hand hygiene skills are regarded as a key contributor to reduce critical infections and healthcare-associated infections. In a clinical setup, at a Neonatal Intensive Care Unit (NICU), the outcome of a multi-modal training initiative was recorded, where objective feedback was provided to the staff. It was hypothesized that staff at the NICU are more sensitive towards applying increased patient safety measures. Outcomes were recorded as the ability to cover all hand surfaces with Alcohol-Based Handrub (ABHR), modelled as a time-series of measurements. The learning ability to rub in with 1.5 mL and with 3 mL was also assessed. As a secondary outcome, handrub consumption and infection numbers were recorded. It has been observed that some staff members were able to quickly learn the proper hand hygiene, even with the limited 1.5 mL, while others were not capable of acquiring the technique even with 3 mL. When analyzing the 1.5 mL group, it was deemed an insufficient ABHR amount, while with 3 mL, the critical necessity of skill training to achieve complete coverage was documented. Identifying these individuals helps the infection control staff to better focus their training efforts. The training led to a 157% increase in handrub consumption. The setting of the study did not allow to show a measurable reduction in the number of hospital infections. It has been concluded that the training method chosen by the staff greatly affects the quality of the outcomes.

Investigation of the Molecular Details of the Interactions of Selenoglycosides and Human Galectin-3.

Human galectin-3 (hGal-3) is involved in a variety of biological processes and is implicated in wide range of diseases. As a result, targeting hGal-3 for clinical applications has become an intense area of research. As a step towards the development of novel hGal-3 inhibitors, we describe a study of the binding of two Se-containing hGal-3 inhibitors, specifically that of di(ß-D-galactopyranosyl)selenide (SeDG), in which two galactose rings are linked by one Se atom and a di(ß-D-galactopyranosyl)diselenide (DSeDG) analogue with a diseleno bond between the two sugar units. The binding affinities of these derivatives to hGal-3 were determined by 15N-1H HSQC NMR spectroscopy and fluorescence anisotropy titrations in solution, indicating a slight decrease in the strength of interaction for SeDG compared to thiodigalactoside (TDG), a well-known inhibitor of hGal-3, while DSeDG displayed a much weaker interaction strength. NMR and FA measurements showed that both seleno derivatives bind to the canonical S face site of hGal-3 and stack against the conserved W181 residue also confirmed by X-ray crystallography, revealing canonical properties of the interaction. The interaction with DSeDG revealed two distinct binding modes in the crystal structure which are in fast exchange on the NMR time scale in solution, explaining a weaker interaction with hGal-3 than SeDG. Using molecular dynamics simulations, we have found that energetic contributions to the binding enthalpies mainly differ in the electrostatic interactions and in polar solvation terms and are responsible for weaker binding of DSeDG compared to SeDG. Selenium-containing carbohydrate inhibitors of hGal-3 showing canonical binding modes offer the potential of becoming novel hydrolytically stable scaffolds for a new class of hGal-3 inhibitors.

Introducing 77Se NMR Spectroscopy to Analyzing Galectin -Ligand Interaction.

Their emerging nature as multifunctional effectors explains the large interest to monitor glycan binding to galectins and to define bound-state conformer(s) of their ligands in solution. Basically, NMR spectroscopy facilitates respective experiments. Towards developing new and even better approaches for these purposes, extending the range of exploitable isotopes beyond 1H, 13C, and 15N offers promising perspectives. Having therefore prepared selenodigalactoside and revealed its bioactivity as galectin ligand, monitoring of its binding by 77Se NMR spectroscopy at a practical level becomes possible by setting up a 2D 1H, 77Se CPMG-HSQBMC experiment including CPMG-INEPT long-range transfer. This first step into applying 77Se as sensor for galectin binding substantiates its potential for screening relative to inhibitory potencies in compound mixtures and for achieving sophisticated epitope mapping. The documented strategic combination of synthetic carbohydrate chemistry and NMR spectroscopy prompts to envision to work with isotopically pure 77Se-containing ß-galactosides and to build on the gained experience with 77Se by adding 19F as second sensor in doubly labeled glycosides.

77Se-Enriched Selenoglycoside Enables Significant Enhancement in NMR Spectroscopic Monitoring of Glycan-Protein Interactions.

Detailed investigation of ligand-protein interactions is essential for better understanding of biological processes at the molecular level. Among these binding interactions, the recognition of glycans by lectins is of particular importance in several diseases, such as cancer; therefore, inhibition of glycan-lectin/galectin interactions represents a promising perspective towards developing therapeutics controlling cancer development. The recent introduction of 77Se NMR spectroscopy for monitoring the binding of a selenoglycoside to galectins prompted interest to optimize the sensitivity by increasing the 77Se content from the natural 7.63% abundance to 99%. Here, we report a convenient synthesis of 77Se-enriched selenodigalactoside (SeDG), which is a potent ligand of the medically relevant human galectin-3 protein, and proof of the expected sensitivity gain in 2D 1H, 77Se correlation NMR experiments. Our work opens perspectives for adding isotopically enriched selenoglycans for rapid monitoring of lectin-binding of selenated as well as non-selenated ligands and for ligand screening in competition experiments.

Selenoglycosides as Lectin Ligands: 77 Se-Edited CPMG-HSQMBC NMR Spectroscopy To Monitor Biomedically Relevant Interactions.

The fundamental importance of protein-glycan recognition calls for specific and sensitive high-resolution techniques for their detailed analysis. After the introduction of 19 F NMR spectroscopy to study the recognition of fluorinated glycans, a new 77 Se NMR spectroscopy method is presented for complementary studies of selenoglycans with optimised resolution and sensitivity, in which direct NMR spectroscopy detection on 77 Se is replaced by its indirect observation in a 2D 1 H,77 Se HSQMBC spectrum. In contrast to OH/F substitution, O/Se exchange allows the glycosidic bond to be targeted. As an example, selenodigalactoside recognition by three human galectins and a plant toxin is readily indicated by signal attenuation and line broadening in the 2D 1 H,77 Se HSQMBC spectrum, in which CPMG-INEPT long-range transfer ensures maximal detection sensitivity, clean signal phases, and reliable ligand ranking. By monitoring competitive displacement of a selenated spy ligand, the selective 77 Se NMR spectroscopy approach may also be used to screen non-selenated compounds. Finally, 1 H,77 Se CPMG-INEPT transfer allows further NMR sensors of molecular interaction to be combined with the specificity and resolution of 77 Se NMR spectroscopy.

Bivalent glycoconjugates based on 1,5-diazabicyclo[3.3.0]octa-3,6-diene-2,8-dione ("bimane") as a central scaffold.

The heteroaromatic fused diazabicyclic "bimane" ring system, discovered four decades ago, is endowed with remarkable chemical and photophysical properties. No carbohydrate derivatives of bimanes have, however, been described thus far. Here we report on the syntheses of a range of bimanes decorated with various glycosyl residues. Mono- and disaccharide residues were attached to syn- or anti-bimane central cores via thio-, disulfido- or selenoglycosidic linkages to obtain novel fluorescent or nonfluorescent glycoconjugates. Cu(I)-catalyzed cycloaddition of glycosyl azides to a bimane diethynyl derivative furnished further bivalent glycoconjugates with sugar residues linked to the central bimane core via 1,2,3-triazole rings. We have determined the crystal and molecular structures of several glycosylated and non-glycosylated bimanes and report fluorescence data for the new compounds.

A Feature Ranking and Selection Algorithm for Brain Tumor Segmentation in Multi-Spectral Magnetic Resonance Image Data.

Accuracy is the most important quality marker in medical image segmentation. However, when the task is to handle large volumes of data, the relevance of processing speed rises. In machine learning solutions the optimization of the feature set can significantly reduce the computational load. This paper presents a method for feature selection and applies it in the context of a brain tumor detection and segmentation problem in multi-spectral magnetic resonance image data. Starting from an initial set of 104 features involved in an existing ensemble learning solution that employs binary decision trees, a reduced set of features is obtained using a iterative algorithm based on a composite criterion. In each iteration, features are ranked according to the frequency of usage and the correctness of the decisions to which they contribute. Lowest ranked features are iteratively eliminated as long as the segmentation accuracy is not damaged. The final reduced set of 13 features provide the same accuracy in the whole tumor segmentation process as the initial one, but three times faster.

Aralkyl selenoglycosides and related selenosugars in acetylated form activate protein phosphatase-1 and -2A.

Aralkyl and aryl selenoglycosides as well as glycosyl selenocarboxylate derivatives were assayed on the activity of protein phosphatase-1 (PP1) and -2A (PP2A) catalytic subunits (PP1c and PP2Ac) in search of compounds for PP1c and PP2Ac effectors. The majority of tested selenoglycosides activated both PP1c and PP2Ac by ∼2-4-fold in a phosphatase assay with phosphorylated myosin light chain substrate when the hydroxyl groups of the glycosyl moiety were acetylated, but they were without any effects in the non-acetylated forms. A peptide from the myosin phosphatase target subunit-1 (MYPT123-38) that included an RVxF PP1c-binding motif attenuated activation of PP1c by 2-Trifluoromethylbenzyl 2,3,4,6-tetra-O-acetyl-1-seleno-ß-d-glucopyranoside (TFM-BASG) and 4-Bromobenzyl 2,3,4,6-tetra-O-acetyl-1-seleno-ß-d-glucopyranoside (Br-BASG). MYPT123-38 stimulated PP2Ac and contributed to PP2Ac activation exerted by either Br-BASG or TFM-BASG. Br-BASG and TFM-BASG suppressed partially binding of PP1c to MYPT1 in surface plasmon resonance based binding experiments. Molecular docking predicted that the hydrophobic binding surfaces in PP1c for interaction with either the RVxF residues of PP1c-interactors or selenoglycosides are partially overlapped. Br-BASG and TFM-BASG caused a moderate increase in the phosphatase activity of HeLa cells in 1 h, and suppressed cell viability in 24 h incubations. In conclusion, our present study identified selenoglycosides as novel activators of PP1 and PP2A as well as provided insights into the structural background of their interactions establishing a molecular model for future design of more efficient phosphatase activator molecules.

Diglycosyl diselenides alter redox homeostasis and glucose consumption of infective African trypanosomes.

With the aim to develop compounds able to target multiple metabolic pathways and, thus, to lower the chances of drug resistance, we investigated the anti-trypanosomal activity and selectivity of a series of symmetric diglycosyl diselenides and disulfides. Of 18 compounds tested the fully acetylated forms of di-ß-D-glucopyranosyl and di-ß-D-galactopyranosyl diselenides (13 and 15, respectively) displayed strong growth inhibition against the bloodstream stage of African trypanosomes (EC50 0.54 µM for 13 and 1.49 µM for 15) although with rather low selectivity (SI < 10 assayed with murine macrophages). Nonacetylated versions of the same sugar diselenides proved to be, however, much less efficient or completely inactive to suppress trypanosome growth. Significantly, the galactosyl (15), and to a minor extent the glucosyl (13), derivative inhibited glucose catabolism but not its uptake. Both compounds induced redox unbalance in the pathogen. In vitro NMR analysis indicated that diglycosyl diselenides react with glutathione, under physiological conditions, via formation of selenenylsulfide bonds. Our results suggest that non-specific cellular targets as well as actors of the glucose and the redox metabolism of the parasite may be affected. These molecules are therefore promising leads for the development of novel multitarget antitrypanosomal agents.

[Hand hygiene technique assessment using electronic equipment in 26 Hungarian healthcare institutions]. / A kézhigiénés technika vizsgálata elektronikus ellenorzo berendezés segítségével 26 magyarországi betegellátó intézményben.

INTRODUCTION: Hand hygiene is probably the most effective tool of nosocomial infection prevention, however, proper feedback and control is needed to develop the individual hand hygiene practice. AIM: Assessing the efficiency of modern education tools, and digital demonstration and verification equipment during their wide-range deployment. METHOD: 1269 healthcare workers took part in a training organized by our team. The training included the assessment of the participants' hand hygiene technique to identify the most often missed areas. The hand hygiene technique was examined by a digital device. RESULTS: 33% of the participants disinfected their hands incorrectly. The most often missed sites are the fingertips (33% on the left hand, 37% on the right hand) and the thumbs (42% on the left hand, 32% on the right hand). CONCLUSION: The feedback has a fundamental role in the development of the hand hygiene technique. With the usage of electronic devices feedback can be provided efficiently and simply. Orv Hetil. 2017; 158(29): 1143-1148.

Increase of Short-Term Heart Rate Variability Induced by Blood Pressure Measurements during Ambulatory Blood Pressure Monitoring.

Objective. The possible effect of blood pressure measurements per se on heart rate variability (HRV) was studied in the setting of concomitant ambulatory blood pressure monitoring (ABPM) and Holter ECG monitoring (HM). Methods. In 25 hypertensive patients (14 women and 11 men, mean age: 58.1 years), 24-hour combined ABPM and HM were performed. For every blood pressure measurement, 2-minute ECG segments (before, during, and after measurement) were analyzed to obtain time domain parameters of HRV: SDNN and rMSSD. Mean of normal RR intervals (MNN), SDNN/MNN, and rMSSD/MNN were calculated, too. Parameter variations related to blood pressure measurements were analyzed using one-way ANOVA with multiple comparisons. Results. 2281 measurements (1518 during the day and 763 during the night) were included in the analysis. Both SDNN and SDNN/MNN had a constant (the same for 24-hour, daytime, and nighttime values) and significant change related to blood pressure measurements: an increase during measurements and a decrease after them (p < 0.01 for any variation). Conclusion. In the setting of combined ABPM and HM, the blood pressure measurement itself produces an increase in short-term heart rate variability. Clarifying the physiological basis and the possible clinical value of this phenomenon needs further studies.

Bivalent O-glycoside mimetics with S/disulfide/Se substitutions and aromatic core: Synthesis, molecular modeling and inhibitory activity on biomedically relevant lectins in assays of increasing physiological relevance.

The emerging significance of recognition of cellular glycans by lectins for diverse aspects of pathophysiology is a strong incentive for considering development of bioactive and non-hydrolyzable glycoside derivatives, for example by introducing S/Se atoms and the disulfide group instead of oxygen into the glycosidic linkage. We report the synthesis of 12 bivalent thio-, disulfido- and selenoglycosides attached to benzene/naphthalene cores. They present galactose, for blocking a plant toxin, or lactose, the canonical ligand of adhesion/growth-regulatory galectins. Modeling reveals unrestrained flexibility and inter-headgroup distances too small to bridge two sites in the same lectin. Inhibitory activity was first detected by solid-phase assays using a surface-presented glycoprotein, with relative activity enhancements per sugar unit relative to free cognate sugar up to nearly 10fold. Inhibitory activity was also seen on lectin binding to surfaces of human carcinoma cells. In order to proceed to characterize this capacity in the tissue context monitoring of lectin binding in the presence of inhibitors was extended to sections of three types of murine organs as models. This procedure proved to be well-suited to determine relative activity levels of the glycocompounds to block binding of the toxin and different human galectins to natural glycoconjugates at different sites in sections. The results on most effective inhibition by two naphthalene-based disulfides and a selenide raise the perspective for broad applicability of the histochemical assay in testing glycoclusters that target biomedically relevant lectins.

A modified two-stage Markov clustering algorithm for large and sparse networks.

BACKGROUND: Graph-based hierarchical clustering algorithms become prohibitively costly in both execution time and storage space, as the number of nodes approaches the order of millions. OBJECTIVE: A fast and highly memory efficient Markov clustering algorithm is proposed to perform the classification of huge sparse networks using an ordinary personal computer. METHODS: Improvements compared to previous versions are achieved through adequately chosen data structures that facilitate the efficient handling of symmetric sparse matrices. Clustering is performed in two stages: the initial connected network is processed in a sparse matrix until it breaks into isolated, small, and relatively dense subgraphs, which are then processed separately until convergence is obtained. An intelligent stopping criterion is also proposed to quit further processing of a subgraph that tends toward completeness with equal edge weights. The main advantage of this algorithm is that the necessary number of iterations is separately decided for each graph node. RESULTS: The proposed algorithm was tested using the SCOP95 and large synthetic protein sequence data sets. The validation process revealed that the proposed method can reduce 3-6 times the processing time of huge sequence networks compared to previous Markov clustering solutions, without losing anything from the partition quality. CONCLUSIONS: A one-million-node and one-billion-edge protein sequence network defined by a BLAST similarity matrix can be processed with an upper-class personal computer in 100 minutes. Further improvement in speed is possible via parallel data processing, while the extension toward several million nodes needs intermediary data storage, for example on solid state drives.

Can Ambulatory Blood Pressure Variability Contribute to Individual Cardiovascular Risk Stratification?

Objective. The aim of this study is to define the normal range for average real variability (ARV) and to establish whether it can be considered as an additional cardiovascular risk factor. Methods. In this observational study, 110 treated hypertensive patients were included and admitted for antihypertensive treatment adjustment. Circadian blood pressure was recorded with validated devices. Blood pressure variability (BPV) was assessed according to the ARV definition. Based on their variability, patients were classified into low, medium, and high variability groups using the fuzzy c-means algorithm. To assess cardiovascular risk, blood samples were collected. Characteristics of the groups were compared by ANOVA tests. Results. Low variability was defined as ARV below 9.8 mmHg (32 patients), medium as 9.8-12.8 mmHg (48 patients), and high variability above 12.8 mmHg (30 patients). Mean systolic blood pressure was 131.2 ± 16.7, 135.0 ± 12.1, and 141.5 ± 11.4 mmHg in the low, medium, and high variability groups, respectively (p = 0.0113). Glomerular filtration rate was 78.6 ± 29.3, 74.8 ± 26.4, and 62.7 ± 23.2 mL/min/1.73 m(2) in the low, medium, and high variability groups, respectively (p = 0.0261). Conclusion. Increased values of average real variability represent an additional cardiovascular risk factor. Therefore, reducing BP variability might be as important as achieving optimal BP levels, but there is need for further studies to define a widely acceptable threshold value.