Dataset Information

EdgeScaping: Mapping the spatial distribution of pairwise gene expression intensities.

ABSTRACT: Gene co-expression networks (GCNs) are constructed from Gene Expression Matrices (GEMs) in a bottom up approach where all gene pairs are tested for correlation within the context of the input sample set. This approach is computationally intensive for many current GEMs and may not be scalable to millions of samples. Further, traditional GCNs do not detect non-linear relationships missed by correlation tests and do not place genetic relationships in a gene expression intensity context. In this report, we propose EdgeScaping, which constructs and analyzes the pairwise gene intensity network in a holistic, top down approach where no edges are filtered. EdgeScaping uses a novel technique to convert traditional pairwise gene expression data to an image based format. This conversion not only performs feature compression, making our algorithm highly scalable, but it also allows for exploring non-linear relationships between genes by leveraging deep learning image analysis algorithms. Using the learned embedded feature space we implement a fast, efficient algorithm to cluster the entire space of gene expression relationships while retaining gene expression intensity. Since EdgeScaping does not eliminate conventionally noisy edges, it extends the identification of co-expression relationships beyond classically correlated edges to facilitate the discovery of novel or unusual expression patterns within the network. We applied EdgeScaping to a human tumor GEM to identify sets of genes that exhibit conventional and non-conventional interdependent non-linear behavior associated with brain specific tumor sub-types that would be eliminated in conventional bottom-up construction of GCNs. Edgescaping source code is available at https://github.com/bhusain/EdgeScaping under the MIT license.

SUBMITTER: Husain B

PROVIDER: S-EPMC6684082 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

EdgeScaping: Mapping the spatial distribution of pairwise gene expression intensities.

Husain Benafsh B Feltus F Alex FA

PloS one 20190806 8

Gene co-expression networks (GCNs) are constructed from Gene Expression Matrices (GEMs) in a bottom up approach where all gene pairs are tested for correlation within the context of the input sample set. This approach is computationally intensive for many current GEMs and may not be scalable to millions of samples. Further, traditional GCNs do not detect non-linear relationships missed by correlation tests and do not place genetic relationships in a gene expression intensity context. In this rep ...[more]

PMID: 31386677

Similar Datasets

Project description:Alzheimer's disease (AD) is a neurodegenerative disorder defined by the progressive formation and spread of fibrillar aggregates of Aβ peptide and tau protein. Polymorphic forms of these aggregates may contribute to disease in varying ways since different neuropathologies appear to be associated with different sets of fibrillar structures and follow distinct pathological trajectories that elicit characteristic clinical phenotypes. The molecular mechanisms underlying the spread of these aggregates in disease may include nucleation, replication, and migration all of which could vary with polymorphic form, stage of disease, and region of brain. Given the linkage between mechanisms of progression and distribution of polymorphs, mapping the distribution of fibrillar structures in situ has the potential to discriminate between mechanisms of progression. However, the means of carrying out this mapping are limited. Optical microscopy lacks the resolution to discriminate between polymorphs in situ, and higher resolution tools such as ssNMR and cryoEM require the isolation of fibrils from tissue, destroying relevant spatial information. Here, we demonstrate the use of scanning x-ray microdiffraction (XMD) to map the locations of fibrillar polymorphs of Aβ peptides and tau protein in histological thin sections of human brain tissue. Coordinated examination of serial sections by immunohistochemistry was used to aid in the interpretation of scattering patterns and to put the observations in a broader anatomical context. Scattering from lesions in tissue shown to be rich in Aβ fibrils by immunohistochemistry exhibited scattering patterns with a prototypical 4.7 Å cross-β peak, and overall intensity distribution that compared well with that predicted from high resolution structures. Scattering from lesions in tissue with extensive tau pathology also exhibited a 4.7 Å cross-β peak but with intensity distributions that were distinct from those seen in Aβ-rich regions. In summary, these observations demonstrate that XMD is a rich source of information on the distribution of fibrillar polymorphs in diseased human brain tissue. When used in coordination with neuropathological examination it has the potential to provide novel insights into the molecular mechanisms underlying disease.

Project description:Parasites, especially brain-encysting trematodes, can have an impact on host behaviour, facilitating the transmission to next host and completion of the life cycle, but insufficient research has been done on whether specific brain regions are targeted. Using Cardiocephaloides longicollis as a laboratory model, the precise distribution of metacercariae in experimentally-infected, wild and farmed fish was mapped. The brain regions targeted by this parasite were explored, also from a histologic perspective, and potential pathogenic effects were evaluated. Experimental infections allowed to reproduce the natural infection intensity of C. longicollis, with four times higher infection intensity at the higher dose (150 vs 50 cercariae). The observed metacercarial distribution, similar among all fish groups, may reflect a trematode species-specific pattern: metacercariae occur with highest density in the optic lobe area (primarily infecting the periventricular gray zone of optic tectum) and the medulla oblongata, whereas other areas such as the olfactory lobes and cerebellar lobes may be occupied when the more frequently invaded parts of the brain were crowded. Mono- and multicysts (i.e. formed either with a single metacercaria, or with 2-25 metacercariae encapsulated together) may be formed depending on the aggregation and timing of metacercariae arrival, with minor host inflammatory response. Larvae of C. longicollis colonizing specific brain areas may have an effect on the functions associated with these areas, which are generally related to sensory and motor functions, but are also related to other host fitness traits such as school maintenance or recognition of predators. The detailed information on the extent and distribution of C. longicollis in fish encephalon sets the ground to understand the effects of brain parasites on fish, but further investigation to establish if C. longicollis, through purely mechanical damage (e.g., occupation, pressure and displacement), has an actual impact on host behaviour remains to be tested under controlled experimental conditions.

Dataset Information

EdgeScaping: Mapping the spatial distribution of pairwise gene expression intensities.

Publications

EdgeScaping: Mapping the spatial distribution of pairwise gene expression intensities.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets