A framework for biomedical figure segmentation towards image-based document retrieval

The figures included in many of the biomedical publications play an important role in understanding the biological experiments and facts described within. Recent studies have shown that it is possible to integrate the information that is extracted from figures in classical document classification and retrieval tasks in order to improve their accuracy. One important observation about the figures included in biomedical publications is that they are often composed of multiple subfigures or panels, each describing different methodologies or results. The use of these multimodal figures is a common practice in bioscience, as experimental results are graphically validated via multiple methodologies or procedures. Thus, for a better use of multimodal figures in document classification or retrieval tasks, as well as for providing the evidence source for derived assertions, it is important to automatically segment multimodal figures into subfigures and panels. This is a challenging task, however, as different panels can contain similar objects (i.e., barcharts and linecharts) with multiple layouts. Also, certain types of biomedical figures are text-heavy (e.g., DNA sequences and protein sequences images) and they differ from traditional images. As a result, classical image segmentation techniques based on low-level image features, such as edges or color, are not directly applicable to robustly partition multimodal figures into single modal panels.In this paper, we describe a robust solution for automatically identifying and segmenting unimodal panels from a multimodal figure. Our framework starts by robustly harvesting figure-caption pairs from biomedical articles. We base our approach on the observation that the document layout can be used to identify encoded figures and figure boundaries within PDF files. Taking into consideration the document layout allows us to correctly extract figures from the PDF document and associate their corresponding caption. We combine pixel-level representations of the extracted images with information gathered from their corresponding captions to estimate the number of panels in the figure. Thus, our approach simultaneously identifies the number of panels and the layout of figures.In order to evaluate the approach described here, we applied our system on documents containing protein-protein interactions (PPIs) and compared the results against a gold standard that was annotated by biologists. Experimental results showed that our automatic figure segmentation approach surpasses pure caption-based and image-based approaches, achieving a 96.64% accuracy. To allow for efficient retrieval of information, as well as to provide the basis for integration into document classification and retrieval systems among other, we further developed a web-based interface that lets users easily retrieve panels containing the terms specified in the user queries.     

Growth temperature acclimation byAmoeba proteus: Effects on cytoplasmic organelle morphology

Summary The effects on subcellular morphology of maintaining amoebae at temperatures other than 20 C (the routine culture temperature) were assessed. Estimations of cycling potential at each temperature confirmed that acclimation had affected gross cell functioning. Generation times ranged from no division at 6 C, to an optimal minimum of 2 days at 22 C.Organelle morphology changes were studied after 5 days of growth at the new temperatures; alterations were most evident at the extremes of 6 and 28 C. The main mitochondrial alteration resulted in changes to the ratio of Type I: Type II organelles; with a decrease in Type I forms away from the optimal range of 20–22 C. Extended culturing at 6 C generated mitochondrial matrical inclusions. Ribosomal attachment to the endoplasmic reticulum, a common feature of 20 C-grown cells, decreased at the temperature extremes, where an increase in free ribosomes occurred. Upon extended culture at 6 C helical structures, usually observed in groups only within the nucleus, were also present in the cytoplasm. Golgi complexes were less common in cells maintained at extreme temperatures and often showed differences in shape. These changes were all reversible on a return to culturing at 20 C.The possible functional significance of these findings is discussed.     

Reactivation of organelle movements along the cytoskeletal framework of a giant freshwater ameba

The peripheral feeding network of the giant freshwater ameba Reticulomyxa can be easily and rapidly lysed to produce an extensive, stable, and completely exposed cytoskeletal framework of colinear microtubules and microfilaments. Most of the organelles that remain attached to this framework resume rapid saltatory movements at rates of up to 20 micron/s if ATP is added. This lysed model system is also capable of other forms of motility, namely an active splaying of microtubule bundles and bulk streaming. Reactivation does not occur with other nucleoside triphosphates, requires Mg ions, is insensitive to even high concentrations of erythro-9-(3-[2-hydroxynonyl]) adenine, is sensitive to vanadate only at concentrations of approximately 100 microM, and is inhibited by N-ethylmaleimide at concentrations greater than 100 microM. The physiology of this reactivation suggests an organelle transport motor distinct from cytoplasmic dynein and possibly the recently described kinesin. This system can serve as a model for elucidating the mechanisms of intracellular transport and, in addition, provides a unique opportunity to examine associations between microtubules and microfilaments.     

Quantitative analysis of organelle abundance, morphology and dynamics

Recent data indicate that morphological characteristics of cell organelles are important for their function in the cell. These characteristics include not only their shape, number and size, but also their distribution in the cell. Moreover, the dynamics of processes that result in changes in these characteristics (e.g. organelle fission, fusion, autophagy, transport) influence the function of the cell. For a better understanding of these processes quantitative approaches are important. Here we give an overview of contemporary biochemical and microscopy methods that are used to quantify organelle abundance, morphology and the kinetics of the processes that cause changes in these properties.     

Effects of organelle shape on fluorescence recovery after photobleaching

The determination of diffusion coefficients from fluorescence recovery data is often complicated by geometric constraints imposed by the complex shapes of intracellular compartments. To address this issue, diffusion of proteins in the lumen of the endoplasmic reticulum (ER) is studied using cell biological and computational methods. Fluorescence recovery after photobleaching (FRAP) experiments are performed in tissue culture cells expressing GFP-KDEL, a soluble, fluorescent protein, in the ER lumen. The three-dimensional (3D) shape of the ER is determined by confocal microscopy and computationally reconstructed. Within these ER geometries diffusion of solutes is simulated using the method of particle strength exchange. The simulations are compared to experimental FRAP curves of GFP-KDEL in the same ER region. Comparisons of simulations in the 3D ER shapes to simulations in open 3D space show that the constraints imposed by the spatial confinement result in two- to fourfold underestimation of the molecular diffusion constant in the ER if the geometry is not taken into account. Using the same molecular diffusion constant in different simulations, the observed speed of fluorescence recovery varies by a factor of 2.5, depending on the particular ER geometry and the location of the bleached area. Organelle shape considerably influences diffusive transport and must be taken into account when relating experimental photobleaching data to molecular diffusion coefficients. This novel methodology combines experimental FRAP curves with high accuracy computer simulations of diffusion in the same ER geometry to determine the molecular diffusion constant of the solute in the particular ER lumen.     

Identification of cancer cell-line origins using fluorescence image-based phenomic screening

Universal phenotyping techniques that can discriminate among various states of biological systems have great potential. We applied 557 fluorescent library compounds to NCI's 60 human cancer cell-lines (NCI-60) to generate a systematic fluorescence phenotypic profiling data. By the kinetic fluorescence intensity analysis, we successfully discriminated the organ origin of all the 60 cell-lines.     

Whole organism high-content screening by label-free, image-based bayesian classification for parasitic diseases

Sole reliance on one drug, Praziquantel, for treatment and control of schistosomiasis raises concerns about development of widespread resistance, prompting renewed interest in the discovery of new anthelmintics. To discover new leads we designed an automated label-free, high content-based, high throughput screen (HTS) to assess drug-induced effects on in vitro cultured larvae (schistosomula) using bright-field imaging. Automatic image analysis and Bayesian prediction models define morphological damage, hit/non-hit prediction and larval phenotype characterization. Motility was also assessed from time-lapse images. In screening a 10,041 compound library the HTS correctly detected 99.8% of the hits scored visually. A proportion of these larval hits were also active in an adult worm ex-vivo screen and are the subject of ongoing studies. The method allows, for the first time, screening of large compound collections against schistosomes and the methods are adaptable to other whole organism and cell-based screening by morphology and motility phenotyping.     

Using the antibody-antigen binding interface to train image-based deep neural networks for antibody-epitope classification

High-throughput B-cell sequencing has opened up new avenues for investigating complex mechanisms underlying our adaptive immune response. These technological advances drive data generation and the need to mine and analyze the information contained in these large datasets, in particular the identification of therapeutic antibodies (Abs) or those associated with disease exposure and protection. Here, we describe our efforts to use artificial intelligence (AI)-based image-analyses for prospective classification of Abs based solely on sequence information. We hypothesized that Abs recognizing the same part of an antigen share a limited set of features at the binding interface, and that the binding site regions of these Abs share share common structure and physicochemical property patterns that can serve as a “fingerprint” to recognize uncharacterized Abs. We combined large-scale sequence-based protein-structure predictions to generate ensembles of 3-D Ab models, reduced the Ab binding interface to a 2-D image (fingerprint), used pre-trained convolutional neural networks to extract features, and trained deep neural networks (DNNs) to classify Abs. We evaluated this approach using Ab sequences derived from human HIV and Ebola viral infections to differentiate between two Abs, Abs belonging to specific B-cell family lineages, and Abs with different epitope preferences. In addition, we explored a different type of DNN method to detect one class of Abs from a larger pool of Abs. Testing on Ab sets that had been kept aside during model training, we achieved average prediction accuracies ranging from 71–96% depending on the complexity of the classification task. The high level of accuracies reached during these classification tests suggests that the DNN models were able to learn a series of structural patterns shared by Abs belonging to the same class. The developed methodology provides a means to apply AI-based image recognition techniques to analyze high-throughput B-cell sequencing datasets (repertoires) for Ab classification.     

CNN and transformer framework for insect pest classification

Insect pests pose a significant and increasing threat to agricultural production worldwide. However, most existing recognition methods are built upon well-known convolutional neural networks, which limits the possibility of improving pest recognition accuracies. This research attempts to overcome this challenge from a novel perspective, constructing a simplified but very useful network for effective insect pest recognition by combining transformer architecture and convolution blocks. First, the representative features are extracted from the input image using a backbone convolutional neural network. Second, a new transformer attention-based classification head is proposed to sufficiently utilize spatial data from the features. With that, we explore different combinations for each module in our model and abstract our model into a simple and scalable architecture; we introduce more effective training strategies, pretrained models and data augmentation methods. Our models performance was evaluated on the IP102 benchmark dataset and achieved classification accuracies of 74.897% and 75.583% with minimal implementation costs at image resolutions of 224 × 224 pixels and 480 × 480 pixels, respectively. Our model also attains accuracies of 99.472% and 97.935% on the D0 dataset and Li's dataset, respectively, with an image resolution of 224 × 224 pixels. The experimental results demonstrate that our method is superior to the state-of-the-art methods on these datasets. Accordingly, the proposed model can be deployed in practice and provides additional insights into the related research.     

Virus morphology: Insights from super-resolution fluorescence microscopy

As epitomised by the COVID-19 pandemic, diseases caused by viruses are one of the greatest health and economic burdens to human society. Viruses are ‘nanostructures’, and their small size (typically less than 200 nm in diameter) can make it challenging to obtain images of their morphology and structure. Recent advances in fluorescence microscopy have given rise to super-resolution techniques, which have enabled the structure of viruses to be visualised directly at a resolution in the order of 20 nm. This mini-review discusses how recent state-of-the-art super-resolution imaging technologies are providing new nanoscale insights into virus structure.     

The association between distribution and octopodid morphology: implications for classification

The Octopodidae occur in virtually all benthic marine habitats; however, species in the family show little overt morphological differentiation. Subfamilies are currently defined by the presence or absence of an ink sac and the number of sucker rows (the presence of an ink sac and a single row of suckers are primitive characters) (Voss, 1988b); subfamily depth ranges arc cited in the diagnoses. Examination of external octopodid morphology through principal components analysis reveals that octopodid morphology correlates with geographic distribution. Low-latitude, shallow-water octopuses typically have narrower bodies and larger suckers on longer arms than do deep sea and high-latitude species. Sucker size inversely correlates with depth distribution, as studies of sucker functional morphology predict (Kier & Smith, 1990). The same characters contribute in a very similar manner to the discrimination of species when grouped by subfamily and when grouped by mean depth distribution. That depth distributions, which correlate with morphology and with the loss of the ink sac, contribute to the definition of these subfamilies, suggests that the subfamilies constitute phenetically similar rather than monophyletic groups. Cladistic analysis is required to reassess octopodid phylogeny.     

Mitochondria isolated in nearly isotonic KCl buffer: Focus on cardiolipin and organelle morphology

Rat liver mitochondria were isolated in parallel in two different isolation buffers: a standard buffer containing mannitol/sucrose and a nearly physiological KCl based solution. The two different organelle preparations were comparatively characterized by respiratory activity, heme content, microsomal and Golgi contamination, electron microscopy and lipid analyses. The substitution of saccharides with KCl in the isolation buffer does not induce the formation of mitoplasts or disruption of mitochondria. Mitochondria isolated in KCl buffer are coupled and able to maintain a stable transmembrane charge separation. A number of biochemical and functional differences between the two organelle preparations are described; in particular KCl mitochondria exhibit lower cardiolipin content and smaller intracristal compartments in comparison with the standard mitochondrial preparation.     

The trilobite family Nileidae: morphology and classification

Species of genera currently referred to Nileidae are reviewed, and those of Hemibarrandia , Lakaspis , Peraspis and Symphysurina are excluded from the family. Nileidae are united in having a distinctive form of the hypostome, the glabellar organ, in the shallowness or absence of external furrows on the axial and pleural regions, and in the development of strong ventral ridges on the axial region. It is contended that the glabellar organ of nileids and illaenids may not be homologous with the glabellar tubercle of asaphids, that the median ventral suture is not exclusively a character of Asaphina, and doubt is cast on the identification of an asaphoid protaspis as being that of Nileus . These arguments provide a case for allying Nileidae with the Illaenidae, rather than with the Asaphina.     

Proteomics of a fuzzy organelle: interphase chromatin

Chromatin proteins mediate replication, regulate expression, and ensure integrity of the genome. So far, a comprehensive inventory of interphase chromatin has not been determined. This is largely due to its heterogeneous and dynamic composition, which makes conclusive biochemical purification difficult, if not impossible. As a fuzzy organelle, it defies classical organellar proteomics and cannot be described by a single and ultimate list of protein components. Instead, we propose a new approach that provides a quantitative assessment of a protein's probability to function in chromatin. We integrate chromatin composition over a range of different biochemical and biological conditions. This resulted in interphase chromatin probabilities for 7635 human proteins, including 1840 previously uncharacterized proteins. We demonstrate the power of our large‐scale data‐driven annotation during the analysis of cyclin‐dependent kinase (CDK) regulation in chromatin. Quantitative protein ontologies may provide a general alternative to list‐based investigations of organelles and complement Gene Ontology.     

The peroxisome: still a mysterious organelle

More than half a century of research on peroxisomes has revealed unique features of this ubiquitous subcellular organelle, which have often been in disagreement with existing dogmas in cell biology. About 50 peroxisomal enzymes have so far been identified, which contribute to several crucial metabolic processes such as β-oxidation of fatty acids, biosynthesis of ether phospholipids and metabolism of reactive oxygen species, and render peroxisomes indispensable for human health and development. It became obvious that peroxisomes are highly dynamic organelles that rapidly assemble, multiply and degrade in response to metabolic needs. However, many aspects of peroxisome biology are still mysterious. This review addresses recent exciting discoveries on the biogenesis, formation and degradation of peroxisomes, on peroxisomal dynamics and division, as well as on the interaction and cross talk of peroxisomes with other subcellular compartments. Furthermore, recent advances on the role of peroxisomes in medicine and in the identification of novel peroxisomal proteins are discussed.     

SSK-DDoS: distributed stream processing framework based classification system for DDoS attacks

Cluster Computing - Distributed denial of service (DDoS) is an immense threat for Internet based-applications and their resources. It immediately floods the victim system by transmitting a large...     

Migrasomes: a new organelle of migrating cells

Cell migration is a multi-step process that involves the coordinated action of signaling networks, cytoskeletal dynamics and vesicular trafficking, leading to protrusion and adhesion at the leading edge of cells and contraction and detachment at their rear. In a recent paper in Cell Research, Ma et al. describe the biogenesis of a new exosome-like organelle — named migrasomes — that derive from retraction fibers at the rear of migrating cells and their potential roles in inter-cellular signaling.Cell migration is a complex process of fundamental importance in numerous physiological or pathological processes, including tissue development, repair and growth, as well as during cancer cell metastasis^1,2,3. Cells use multiple modes of migration depending on the properties of the extracellular matrix (ECM) upon which they move, for example its architecture (two or three dimensional), stiffness and composition, and also on cellular determinants, such as cell-cell and cell-ECM adhesiveness, traction forces and proteolysis⁴. However, in general cell migration is a cyclical process involving: 1) lamellipodial protrusion at the leading edge, 2) cell adhesion by integrin receptors, 3) creation of cell tension, 4) cell contraction, and 5) retraction of the cell rear end. Accomplishing these steps requires integration of signaling, cytoskeletal dynamics, adhesion and membrane trafficking^1,2,3,4.Although most research on migration focuses on events at the cell''s leading edge, the rear end (trailing edge) is equally important because without the coordinated disassembly of cell adhesions and the recycling of ECM receptors, cells would not move⁵. Long tubular cytoplasmic expansions called retraction fibers (fibrils) are observed at the rear end of forward migrating cells⁶, and also during abrupt detachment from substrates⁷. At the tips of these thin extensions, membrane shedding can occur, leaving integrin-containing vesicles attached to the substrate along the migratory path^7,8. Previously thought to be passively deposited cell remnants, Ma et al.⁹ provide evidence for the active transport of materials from the cell body to a specialized subcompartment within these fibers, called migrasomes. As the cell moves and the fibers retract, the migrasomes are released to function as a potentially new and structurally distinct category of exosome-like vesicles that transmit signals between migrating cells (Figure 1).Open in a separate windowFigure 1Migrasomes form within retraction fibers emerging from the trailing edge of migrating cells. They are released and deposited on the substrate as the fibers collapse. Migrasomes may represent a vector for inter-cellular communication as they can be engulfed by trailing cells, providing guidance cues or other information.Exosomes are micro-vesicles of endosomal origin thought to be released after fusion of multivesicular endosomes (MVBs) with the plasma membrane¹⁰. Exosomes, which contain cytosolic contents including miRNAs, as well as lipids and membrane receptors, are released from one cell and believed to bind and/or be taken up by another, providing a means of inter-cellular communication. Interestingly, exosomes have been shown to induce cell migration and invasion in cancer models¹¹. Proteomic analysis of purified migrasomes suggests that they are compositionally related to exosomes⁹, and in particular are enriched in tetraspanins, membrane protein markers of MVBs. However, unlike exosomes, the released migrasomes are large in diameter (0.5-1.2 μm) and contain variable numbers of small internal vesicles.To gain more insights into migrasome biogenesis, Ma et al. expressed and tracked GFP-tagged tetraspanin-4 (TSPAN4-GFP), a protein previously localized to retraction fibers⁷ and shown to be enriched in isolated migrasomes. By time-lapse microscopy of migrating cells⁹, migrasomes form as bulb-like structures at the tips of, or at the intersections between, retraction fibers. Importantly, these bulbs continue to grow in diameter, receiving cytosolic input continuously over an ∼2 h period as the migrating cell advances. When the retraction fibers break, migrasomes are released as a package of vesicles enclosed within a single limiting membrane. Eventually the migrasome dissociates from the substrate and is released into the medium, unless it is engulfed by an oncoming cell (see below).Not surprisingly, given their relationship with retraction fibers, migrasome formation is dependent on cell migration. Thus, the number of migrasomes formed per cell is increased when migration is enhanced (e.g., by coating surfaces with fibronectin, which stimulates cell adhesion and migration¹² or by knocking down SHARPIN, an endogenous inhibitor of β1 integrin activation¹³) and decreased when migration is inhibited (e.g., by treating cells with blebbistatin, a myosin II blocker¹⁴). Importantly, given cell type differences in migration and differences in migration on 2D vs 3D matrices and in vivo, the authors used TSPAN4-GFP transfection and scanning electron microscopy to identify migrasomes formed by multiple cell lines and under multiple conditions. They also identified externally deposited MVBs reminiscent of migrasomes by transmission electron microscopy in various mouse tissues, providing strong evidence for migrasome deposition under physiologically relevant conditions. Although morphologically reminiscent of MVBs, electron micrographs of migrasomes growing within retraction fibers show a collection of free vesicles and cytosolic content and not the deposition of intact MVBs. When the retraction fiber breaks, the released migrasome is surrounded by a single membrane, presumably derived from the surface membrane of the retraction fiber. This is important because the outer membrane of an MVB, unlike the retraction fiber membrane, would be topologically reversed exposing the cytoplasmic surface to the extracellular media.Migrasomes represent a distinct type of extracellular vesicle deposited through a unique mechanism and left in the tracks of migrating cells. Clearly, further work is necessary both to define the cellular components and potential signaling molecules that accumulate in growing bulbs along retraction fibers for release in migrasomes, and to identify the intracellular targeting signals and mechanisms that deposit them there. Most important is to understand the function of migrasomes. One clue comes from striking time-lapse videos showing that other cells following in the path of their depositors can take up migrasomes⁹. Which signals are activated upon contact with and/or internalization of deposited migrasomes? What inter-cellular messages might they be delivering? Could they serve as migratory benchmarks or guidance cues? The discovery of these extracellular vesicles raises many new questions and opens many avenues for future research.