network topology
Dotaz
Zobrazit nápovědu
Optimization of neural network topology, weights and neuron transfer functions for given data set and problem is not an easy task. In this article, we focus primarily on building optimal feed-forward neural network classifier for i.i.d. data sets. We apply meta-learning principles to the neural network structure and function optimization. We show that diversity promotion, ensembling, self-organization and induction are beneficial for the problem. We combine several different neuron types trained by various optimization algorithms to build a supervised feed-forward neural network called Group of Adaptive Models Evolution (GAME). The approach was tested on a large number of benchmark data sets. The experiments show that the combination of different optimization algorithms in the network is the best choice when the performance is averaged over several real-world problems.
One of the aims of high-throughput gene/protein profiling experiments is the identification of biological processes altered between two or more conditions. Pathway analysis is an umbrella term for a multitude of computational approaches used for this purpose. While in the beginning pathway analysis relied on enrichment-based approaches, a newer generation of methods is now available, exploiting pathway topologies in addition to gene/protein expression levels. However, little effort has been invested in their critical assessment with respect to their performance in different experimental setups. Here, we assessed the performance of seven representative methods identifying differentially expressed pathways between two groups of interest based on gene expression data with prior knowledge of pathway topologies: SPIA, PRS, CePa, TAPPA, TopologyGSA, Clipper and DEGraph. We performed a number of controlled experiments that investigated their sensitivity to sample and pathway size, threshold-based filtering of differentially expressed genes, ability to detect target pathways, ability to exploit the topological information and the sensitivity to different pre-processing strategies. We also verified type I error rates and described the influence of overexpression of single genes, gene sets and topological motifs of various sizes on the detection of a pathway as differentially expressed. The results of our experiments demonstrate a wide variability of the tested methods. We provide a set of recommendations for an informed selection of the proper method for a given data analysis task.
- MeSH
- databáze genetické MeSH
- datové soubory jako téma MeSH
- lidé MeSH
- metabolické sítě a dráhy * MeSH
- stanovení celkové genové exprese metody MeSH
- vysoce účinné nukleotidové sekvenování MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
- srovnávací studie MeSH
BACKGROUND: Pathway analysis methods, in which differentially expressed genes are mapped to databases of reference pathways and relative enrichment is assessed, help investigators to propose biologically relevant hypotheses. The last generation of pathway analysis methods takes into account the topological structure of a pathway, which helps to increase both specificity and sensitivity of the findings. Simultaneously, the RNA-Seq technology is gaining popularity and becomes widely used for gene expression profiling. Unfortunately, majority of topological pathway analysis methods remains without implementation and if an implementation exists, it is limited in various factors. RESULTS: We developed a new R/Bioconductor package ToPASeq offering uniform interface to seven distinct topology-based pathway analysis methods, of which three we implemented de-novo and four were adjusted from existing implementations. Apart this, ToPASeq offers a set of tailored visualization functions and functions for importing and manipulating pathways and their topologies, facilitating the application of the methods on different species. The package can be used to compare the differential expression of pathways between two conditions on both gene expression microarray and RNA-Seq data. The package is written in R and is available from Bioconductor 3.2 using AGPL-3 license. CONCLUSION: ToPASeq is a novel package that offers seven distinct methods for topology-based pathway analysis, which are easily applicable on microarray as well as RNA-Seq data, both in human and other species. At the same time, it provides specific tools for visualization of the results.
- MeSH
- genové regulační sítě * MeSH
- lidé MeSH
- počítačová grafika * MeSH
- RNA genetika MeSH
- sekvenční analýza RNA metody MeSH
- signální transdukce * MeSH
- software * MeSH
- stanovení celkové genové exprese metody MeSH
- vysoce účinné nukleotidové sekvenování metody MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
Impairment in spatial navigation (SN) and structural network topology is not limited to patients with Alzheimer's disease (AD) dementia and can be detected earlier in patients with mild cognitive impairment (MCI). We recruited 32 MCI patients (65.91 ± 11.33 years old) and 28 normal cognition patients (NC; 69.68 ± 10.79 years old), all of whom underwent a computer-based battery of SN tests evaluating egocentric, allocentric, and mixed SN strategies and diffusion-weighted and T1-weighted Magnetic Resonance Imaging (MRI). To evaluate the topological features of the structural connectivity network, we calculated its measures such as the global efficiency, local efficiency, clustering coefficient, and shortest path length with GRETNA. We determined the correlation between SN accuracy and network topological properties. Compared to NC, MCI subjects demonstrated a lower egocentric navigation accuracy. Compared with NC, MCI subjects showed significantly decreased clustering coefficients in the left middle frontal gyrus, right rectus, right superior parietal gyrus, and right inferior parietal gyrus and decreased shortest path length in the left paracentral lobule. We observed significant positive correlations of the shortest path length in the left paracentral lobule with both the mixed allocentric-egocentric and the allocentric accuracy measured by the average total errors. A decreased clustering coefficient in the right inferior parietal gyrus was associated with a larger allocentric navigation error. White matter hyperintensities (WMH) did not affect the correlation between network properties and SN accuracy. This study demonstrated that structural connectivity network abnormalities, especially in the frontal and parietal gyri, are associated with a lower SN accuracy, independently of WMH, providing a new insight into the brain mechanisms associated with SN impairment in MCI.
- Publikační typ
- časopisecké články MeSH
The degree of response to subthalamic nucleus deep brain stimulation (STN-DBS) is individual and hardly predictable. We hypothesized that DBS-related changes in cortical network organization are related to the clinical effect. Network analysis based on graph theory was used to evaluate the high-density electroencephalography (HDEEG) recorded during a visual three-stimuli paradigm in 32 Parkinson's disease (PD) patients treated by STN-DBS in stimulation "off" and "on" states. Preprocessed scalp data were reconstructed into the source space and correlated to the behavioral parameters. In the majority of patients (n = 26), STN-DBS did not lead to changes in global network organization in large-scale brain networks. In a subgroup of suboptimal responders (n = 6), identified according to reaction times (RT) and clinical parameters (lower Unified Parkinson's Disease Rating Scale [UPDRS] score improvement after DBS and worse performance in memory tests), decreased global connectivity in the 1-8 Hz frequency range and regional node strength in frontal areas were detected. The important role of the supplementary motor area for the optimal DBS response was demonstrated by the increased node strength and eigenvector centrality in good responders. This response was missing in the suboptimal responders. Cortical topologic architecture is modified by the response to STN-DBS leading to a dysfunction of the large-scale networks in suboptimal responders.
- MeSH
- elektroencefalografie MeSH
- hluboká mozková stimulace * MeSH
- hodnocení výsledků zdravotní péče MeSH
- lidé středního věku MeSH
- lidé MeSH
- mozková kůra patofyziologie MeSH
- nervová síť patofyziologie MeSH
- nucleus subthalamicus patofyziologie MeSH
- Parkinsonova nemoc patofyziologie terapie MeSH
- psychomotorický výkon fyziologie MeSH
- senioři MeSH
- Check Tag
- lidé středního věku MeSH
- lidé MeSH
- mužské pohlaví MeSH
- senioři MeSH
- ženské pohlaví MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
The Drosophila retinal determination gene network occurs in animals generally as a Pax-Six-Eyes absent-Dachshund network (PSEDN). For amphioxus, we describe the complete network of nine PSEDN genes, four of which-AmphiSix1/2, AmphiSix4/5, AmphSix3/6, and AmphiEya-are characterized here for the first time. For amphioxus, in vitro interactions among the genes and proteins of the network resemble those of other animals, except for the absence of Dach-Eya binding. Amphioxus PSEDN genes are expressed in highly stage- and tissue-specific patterns (sometimes conspicuously correlated with the local intensity of cell proliferation) in the gastrular organizer, notochord, somites, anterior central nervous system, peripheral nervous system, pharyngeal endoderm, and the likely homolog of the vertebrate adenohypophysis. In this last tissue, the anterior region expresses all three amphioxus Six genes and is a zone of active cell proliferation, while the posterior region expresses only AmphiPax6 and is non-proliferative. In summary, the topologies of animal PSEDNs, although considerably more variable than originally proposed, are conserved enough to be recognizable among species and among developing tissues; this conservation may reflect indispensable involvement of PSEDNs during the critically important early phases of embryology (e.g. in the control of mitosis, apoptosis, and cell/tissue motility).
- MeSH
- adenohypofýza metabolismus růst a vývoj MeSH
- aktivní transport - buněčné jádro MeSH
- centrální nervový systém metabolismus růst a vývoj MeSH
- Chordata genetika růst a vývoj MeSH
- epidermální buňky MeSH
- epidermis růst a vývoj MeSH
- exprese genu MeSH
- financování organizované MeSH
- fylogeneze MeSH
- homeodoménové proteiny analýza genetika metabolismus MeSH
- ledviny metabolismus růst a vývoj MeSH
- messenger RNA analýza metabolismus MeSH
- neurony aferentní metabolismus MeSH
- proteiny nervové tkáně analýza genetika metabolismus MeSH
- zvířata MeSH
- Check Tag
- zvířata MeSH
Synfire rings are neural circuits capable of conveying synchronous, temporally precise and self-sustained activities in a robust manner. We propose a cell assembly based paradigm for abstract neural computation centered on the concept of synfire rings. More precisely, we empirically show that Hodgkin-Huxley neural networks modularly composed of synfire rings are automata complete. We provide an algorithmic construction which, starting from any given finite state automaton, builds a corresponding Hodgkin-Huxley neural network modularly composed of synfire rings and capable of simulating it. We illustrate the correctness of the construction on two specific examples. We further analyze the stability and robustness of the construction as a function of changes in the ring topologies as well as with respect to cell death and synaptic failure mechanisms, respectively. These results establish the possibility of achieving abstract computation with bio-inspired neural networks. They might constitute a theoretical ground for the realization of biological neural computers.
- MeSH
- akční potenciály fyziologie MeSH
- lidé MeSH
- modely neurologické * MeSH
- neuronové sítě * MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- časopisecké články MeSH
Recognition memory is the ability to recognize previously encountered objects. Even this relatively simple, yet extremely fast, ability requires the coordinated activity of large-scale brain networks. However, little is known about the sub-second dynamics of these networks. The majority of current studies into large-scale network dynamics is primarily based on imaging techniques suffering from either poor temporal or spatial resolution. We investigated the dynamics of large-scale functional brain networks underlying recognition memory at the millisecond scale. Specifically, we analyzed dynamic effective connectivity from intracranial electroencephalography while epileptic subjects (n = 18) performed a fast visual recognition memory task. Our data-driven investigation using Granger causality and the analysis of communities with the Louvain algorithm spotlighted a dynamic interplay of two large-scale networks associated with successful recognition. The first network involved the right visual ventral stream and bilateral frontal regions. It was characterized by early, predominantly bottom-up information flow peaking at 115 ms. It was followed by the involvement of another network with predominantly top-down connectivity peaking at 220 ms, mainly in the left anterior hemisphere. The transition between these two networks was associated with changes in network topology, evolving from a more segregated to a more integrated state. These results highlight that distinct large-scale brain networks involved in visual recognition memory unfold early and quickly, within the first 300 ms after stimulus onset. Our study extends the current understanding of the rapid network changes during rapid cognitive processes.
Computer assisted image acquisition techniques, including confocal microscopy, require efficient tools for an automatic sorting of vast amount of generated image data. The complexity of the classification process, absence of adequate tools, and insufficient amount of reference data has made the automated processing of images challenging. Mastering of this issue would allow implementation of statistical analysis in research areas such as in research on formation of t-tubules in cardiac myocytes. We developed a system aimed at automatic assessment of cardiomyocyte development stages (SAACS). The system classifies confocal images of cardiomyocytes with fluorescent dye stained sarcolemma. We based SAACS on a densely connected convolutional network (DenseNet) topology. We created a set of labelled source images, proposed an appropriate data augmentation technique and designed a class probability graph. We showed that the DenseNet topology, in combination with the augmentation technique is suitable for the given task, and that high-resolution images are instrumental for image categorization. SAACS, in combination with the automatic high-throughput confocal imaging, will allow application of statistical analysis in the research of the tubular system development or remodelling and loss.
- MeSH
- buněčná diferenciace MeSH
- fluorescenční barviva MeSH
- kardiomyocyty cytologie ultrastruktura MeSH
- konfokální mikroskopie metody MeSH
- krysa rodu rattus MeSH
- modely kardiovaskulární MeSH
- neuronové sítě MeSH
- počítačové zpracování obrazu metody MeSH
- sarkolema ultrastruktura MeSH
- strojové učení MeSH
- umělá inteligence MeSH
- zvířata MeSH
- Check Tag
- krysa rodu rattus MeSH
- zvířata MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
