Automated interpretation of subcellular patterns in fluorescence microscope images for location proteomics.

Proteomics, the large scale identification and characterization of many or all proteins expressed in a given cell type, has become a major area of biological research. In addition to information on protein sequence, structure and expression levels, knowledge of a protein's subcellular location is essential to a complete understanding of its functions. Currently, subcellular location patterns are routinely determined by visual inspection of fluorescence microscope images. We review here research aimed at creating systems for automated, systematic determination of location. These employ numerical feature extraction from images, feature reduction to identify the most useful features, and various supervised learning (classification) and unsupervised learning (clustering) methods. These methods have been shown to perform significantly better than human interpretation of the same images. When coupled with technologies for tagging large numbers of proteins and high-throughput microscope systems, the computational methods reviewed here enable the new subfield of location proteomics. This subfield will make critical contributions in two related areas. First, it will provide structured, high-resolution information on location to enable Systems Biology efforts to simulate cell behavior from the gene level on up. Second, it will provide tools for Cytomics projects aimed at characterizing the behaviors of all cell types before, during, and after the onset of various diseases.     

A multiresolution approach to automated classification of protein subcellular location images

Background  

Fluorescence microscopy is widely used to determine the subcellular location of proteins. Efforts to determine location on a proteome-wide basis create a need for automated methods to analyze the resulting images. Over the past ten years, the feasibility of using machine learning methods to recognize all major subcellular location patterns has been convincingly demonstrated, using diverse feature sets and classifiers. On a well-studied data set of 2D HeLa single-cell images, the best performance to date, 91.5%, was obtained by including a set of multiresolution features. This demonstrates the value of multiresolution approaches to this important problem.     

A graphical model approach to automated classification of protein subcellular location patterns in multi-cell images

Background  

Knowledge of the subcellular location of a protein is critical to understanding how that protein works in a cell. This location is frequently determined by the interpretation of fluorescence microscope images. In recent years, automated systems have been developed for consistent and objective interpretation of such images so that the protein pattern in a single cell can be assigned to a known location category. While these systems perform with nearly perfect accuracy for single cell images of all major subcellular structures, their ability to distinguish subpatterns of an organelle (such as two Golgi proteins) is not perfect. Our goal in the work described here was to improve the ability of an automated system to decide which of two similar patterns is present in a field of cells by considering more than one cell at a time. Since cells displaying the same location pattern are often clustered together, considering multiple cells may be expected to improve discrimination between similar patterns.     

Segmentation of color images from serous cytology for automated cell classification

OBJECTIVE: To design an automated system for the classification of cells based on analysis of serous cytology, with the aim of segmenting both cytoplasm and nucleus using color information from the images as the main characteristic of the cells. STUDY DESIGN: The segmentation strategy uses color information coupled with mathematical morphology tools, such as watersheds. Cytoplasm and nuclei of all diagnostic cells are retained; erythrocytes and debris are eliminated. Special techniques are used for the separation of clustered cells. RESULTS: A large set of cells was assessed by experts to score the segmentation success rate. All cells were segmented whatever their spatial configurations. The average success rate was 92.5% for nuclei and 91.1% for cytoplasm. CONCLUSION: This color information-based segmentation of images of serous cells is accurate and provides a useful tool. This segmentation strategy will improve the automated classification of cells.     

Evaluation of methods for detection of fluorescence labeled subcellular objects in microscope images

Background  

Several algorithms have been proposed for detecting fluorescently labeled subcellular objects in microscope images. Many of these algorithms have been designed for specific tasks and validated with limited image data. But despite the potential of using extensive comparisons between algorithms to provide useful information to guide method selection and thus more accurate results, relatively few studies have been performed.     

Software for quantification of labeled bacteria from digital microscope images by automated image analysis

Automated image analysis software, CellC, was developed and validated for quantification of bacterial cells from digital microscope images. CellC enables automated enumeration of bacterial cells, comparison of total count and specific count images [e.g., 4',6-diamino-2-phenylindole (DAPI) and fluorescence in situ hybridization (FISH) images], and provides quantitative estimates of cell morphology. The software includes an intuitive graphical user interface that enables easy usage as well as sequential analysis of multiple images without user intervention. Validation of enumeration reveals correlation to be better than 0.98 when total bacterial counts by CellC are compared with manual enumeration, with all validated image types. The software is freely available and modifiable: the executable files and MATLAB source codes can be obtained at www. cs. tut.fi/sgn/csb/cellc.     

Automatic signal classification in fluorescence in situ hybridization images

BACKGROUND: Previous systems for dot (signal) counting in fluorescence in situ hybridization (FISH) images have relied on an auto-focusing method for obtaining a clearly defined image. Because signals are distributed in three dimensions within the nucleus and artifacts such as debris and background fluorescence can attract the focusing method, valid signals can be left unfocused or unseen. This leads to dot counting errors, which increase with the number of probes. METHODS: The approach described here dispenses with auto-focusing, and instead relies on a neural network (NN) classifier that discriminates between in and out-of-focus images taken at different focal planes of the same field of view. Discrimination is performed by the NN, which classifies signals of each image as valid data or artifacts (due to out of focusing). The image that contains no artifacts is the in-focus image selected for dot count proportion estimation. RESULTS: Using an NN classifier and a set of features to represent signals improves upon previous discrimination schemes that are based on nonadaptable decision boundaries and single-feature signal representation. Moreover, the classifier is not limited by the number of probes. Three classification strategies, two of them hierarchical, have been examined and found to achieve each between 83% and 87% accuracy on unseen data. Screening, while performing dot counting, of in and out-of-focus images based on signal classification suggests an accurate and efficient alternative to that obtained using an auto-focusing mechanism.     

Boosting for tumor classification with gene expression data

MOTIVATION: Microarray experiments generate large datasets with expression values for thousands of genes but not more than a few dozens of samples. Accurate supervised classification of tissue samples in such high-dimensional problems is difficult but often crucial for successful diagnosis and treatment. A promising way to meet this challenge is by using boosting in conjunction with decision trees. RESULTS: We demonstrate that the generic boosting algorithm needs some modification to become an accurate classifier in the context of gene expression data. In particular, we present a feature preselection method, a more robust boosting procedure and a new approach for multi-categorical problems. This allows for slight to drastic increase in performance and yields competitive results on several publicly available datasets. AVAILABILITY: Software for the modified boosting algorithms as well as for decision trees is available for free in R at http://stat.ethz.ch/~dettling/boosting.html.     

High-throughput automated platform for nuclear magnetic resonance-based structural proteomics

The development of new systems and strategies capable of synthesizing any desired soluble, labeled protein or protein fragment on a preparative scale is one of the most important tasks in biotechnology today. The Center for Eukaryotic Structural Genomics (WI, USA), in co-operation with Ehime University (Matsuyama, Japan) and CellFree Sciences Co., Ltd, has developed an automated platform for nuclear magnetic resonance-based structural proteomics that employs wheat germ extracts for cell-free production of labeled protein. The platform utilizes a single construct for all targets without any redesign of the DNA or RNA. Therefore, it offers advantages over commercial cell-free methods utilizing Escherichia coli extracts that require multiple constructs or redesign of the open reading frame. The protein production and labeling protocol is no more costly than E. coli cell-based approaches, is robust and scalable for high-throughput applications. This protocol has been used in the authors center to screen eukaryotic open reading frames from the Arabidopsis thaliana and human genomes and for the determination of nuclear magnetic resonance structures. With the recent addition of the GeneDecoder 1000 (CellFree Sciences Co., Ltd) robotic system, the Center for Eukaryotic Structural Genomics is able to carry out as many as 384 small-scale (50 microl) screening reactions per week. Furthermore, the Protemist (CellFree Sciences Co., Ltd) robotic system enables the Center for Eukaryotic Structural Genomics to carry out 16 production-scale (4 ml) reactions per week. Utilization of this automated platform technology to screen targets for expression and solubility and to produce stable isotope-labeled samples for nuclear magnetic resonance structure determinations is discussed.     

Improved recognition of figures containing fluorescence microscope images in online journal articles using graphical models

MOTIVATION: There is extensive interest in automating the collection, organization and analysis of biological data. Data in the form of images in online literature present special challenges for such efforts. The first steps in understanding the contents of a figure are decomposing it into panels and determining the type of each panel. In biological literature, panel types include many kinds of images collected by different techniques, such as photographs of gels or images from microscopes. We have previously described the SLIF system (http://slif.cbi.cmu.edu) that identifies panels containing fluorescence microscope images among figures in online journal articles as a prelude to further analysis of the subcellular patterns in such images. This system contains a pretrained classifier that uses image features to assign a type (class) to each separate panel. However, the types of panels in a figure are often correlated, so that we can consider the class of a panel to be dependent not only on its own features but also on the types of the other panels in a figure. RESULTS: In this article, we introduce the use of a type of probabilistic graphical model, a factor graph, to represent the structured information about the images in a figure, and permit more robust and accurate inference about their types. We obtain significant improvement over results for considering panels separately. AVAILABILITY: The code and data used for the experiments described here are available from http://murphylab.web.cmu.edu/software.     

A fluorescence microscope attachment for flow-through cytofluorometry

A flow-through system for rapid automated analysis of intracellular fluorescence parameters is described. In combination with commercially available instrumentation for fluorescence microscopy, fluorometry and pulse height analysis, the hydrodynamic focusing flow-through principle can be successfully applied. The specific advantages and disadvantages of this principle compared with the mechanical focusing system are discussed.     

Automated three-dimensional detection and shape classification of dendritic spines from fluorescence microscopy images

A fundamental challenge in understanding how dendritic spine morphology controls learning and memory has been quantifying three-dimensional (3D) spine shapes with sufficient precision to distinguish morphologic types, and sufficient throughput for robust statistical analysis. The necessity to analyze large volumetric data sets accurately, efficiently, and in true 3D has been a major bottleneck in deriving reliable relationships between altered neuronal function and changes in spine morphology. We introduce a novel system for automated detection, shape analysis and classification of dendritic spines from laser scanning microscopy (LSM) images that directly addresses these limitations. The system is more accurate, and at least an order of magnitude faster, than existing technologies. By operating fully in 3D the algorithm resolves spines that are undetectable with standard two-dimensional (2D) tools. Adaptive local thresholding, voxel clustering and Rayburst Sampling generate a profile of diameter estimates used to classify spines into morphologic types, while minimizing optical smear and quantization artifacts. The technique opens new horizons on the objective evaluation of spine changes with synaptic plasticity, normal development and aging, and with neurodegenerative disorders that impair cognitive function.     

A classification model for the Leiden proteomics competition

A strategy is presented to build a discrimination model in proteomics studies. The model is built using cross-validation. This cross-validation step can simply be combined with a variable selection method, called rank products. The strategy is especially suitable for the low-samples-to-variables-ratio (undersampling) case, as is often encountered in proteomics and metabolomics studies. As a classification method, Principal Component Discriminant Analysis is used; however, the methodology can be used with any classifier. A data set containing serum samples from breast cancer patients and healthy controls is analysed. Double cross-validation shows that the sensitivity of the model is 82% and the specificity 86%. Potential putative biomarkers are identified using the variable selection method. In each cross-validation loop a classification model is built. The final classification uses a majority voting scheme from the ensemble classifier.     

高精度相对和绝对定量的等量异位标签在定量蛋白质组学研究中的新进展

蛋白质组学逐渐从定性研究转向定量研究。在定量蛋白质组学技术中,相对和绝对定量的等量异位标签(Isobaric tags for relative and absolute quantitation,iTRAQ)是应用最广泛的技术之一,具有通量高、稳定性强及不受样品来源制约等优点,几乎可以对任意样品进行标记,而且可以同时对多达8个样品进行定量分析,有效地提高了通量。iTRAQ技术不断改进,其定量准确性显著提高,适用的平台越来越多,为微生物、动物、植物、生物医学领域蛋白质及其翻译后修饰组研究创造了条件。文中综述了高精度iTRAQ技术在定量蛋白质组学研究中的最新发展及其应用。     

Image processing for automated erythrocyte classification.

Digital image processing and pattern recognition techniques were applied to determine the feasibility of a natural n-space subgrouping of normal and abnormal peripheral blood erythrocytes into well separated categories. The data consisted of 325 digitized red cells from 11 different cell classes. The analysis resulted in five features: (a) size, (b) roundness, (c) spicularity, (d) eccentricity and (e) central gray level distribution. These features separated the data into six distinct condensed subgroups of red cells. Each subgroup consisted of morphologically similar cells: (a) macrocytes, (b) normocytes, (c) schistocytes, acanthocytes and burr cells, (d) microcytes and spherocytes, (e) elliptocytes, sickle cells and pencil forms and (f) target cells. The concept of a quantitative "red cell differential" was introduced, utilizing these subgroup definitions to establish subpopulations of red cells, with quantifiable indices for the diagnosis of anemia, at the specimen level.     

Design for a fast fluorescence laser scanning microscope

The design of a fast fluorescence laser scanning microscope is described and illustrated, with discussion of the design consideration of the principal components, including the optical elements. The system, now under construction at the Optical Sciences Center of the University of Arizona, is expected to provide very-high-speed scanning, at a high spatial sampling density, of large object areas while retaining a flexibility of applications. The projected scanning rate approximates the rate achieved by flow cytometry; the projected rates of information generation should be orders of magnitude higher.     

Two-dimensional fluorescence difference gel electrophoresis for comparative proteomics profiling

Quantitative proteomics is the workhorse of the modern proteomics initiative. The gel-based and MuDPIT approaches have facilitated vital advances in the measurement of protein expression alterations in normal and disease phenotypic states. The methodological advance in two-dimensional gel electrophoresis (2DGE) has been the multiplexing fluorescent two-dimensional fluorescence difference gel electrophoresis (2D-DIGE). 2D-DIGE is based on direct labeling of lysine groups on proteins with cyanine CyDye DIGE Fluor minimal dyes before isoelectric focusing, enabling the labeling of 2-3 samples with different dyes and electrophoresis of all the samples on the same 2D gel. This capability minimizes spot pattern variability and the number of gels in an experiment while providing simple, accurate and reproducible spot matching. This protocol can be completed in 3-5 weeks depending on the sample size of the experiment and the level of expertise of the investigator.     

The aperture-defined microvolume (ADM) method: automated measurements of enzyme activity using an inverted fluorescence microscope

Summary The aperture-defined microvolume (ADM) method is based on the relatively constant absorbance or fluorescence of a microvolume of homogeneously coloured material, which is defined by the numerical aperture of the objective.This paper describes the princile of the method and discusses the equipment needed. The main applications reported so far for the measurement of enzyme activity are reviewed. Among these are the quantification of ELISA and DASS tests used in immunology, kinetic studies of enzymes in solution using fluorogenic substrates, and the measurement of enzyme activity in single cells or cell fractions that have been isolated by flow sorting.Typical characteristics of automated ADM measurements include a coefficient of variation of less than 3%, a lower detection limit of a few nanogrammes of fluorescing dye (e.g. 4-methylumbelliferone) and a linear relationship between fluorescence yield and fluorophore concentration over a range of 0.01 to 2.5 nmol. The scanning of Terasaki-type trays and 96-well microtitration plates can be completely automated and requires approximately one minute.     

Miniaturized integration of a fluorescence microscope

The light microscope is traditionally an instrument of substantial size and expense. Its miniaturized integration would enable many new applications based on mass-producible, tiny microscopes. Key prospective usages include brain imaging in behaving animals for relating cellular dynamics to animal behavior. Here we introduce a miniature (1.9 g) integrated fluorescence microscope made from mass-producible parts, including a semiconductor light source and sensor. This device enables high-speed cellular imaging across ～0.5 mm2 areas in active mice. This capability allowed concurrent tracking of Ca2+ spiking in >200 Purkinje neurons across nine cerebellar microzones. During mouse locomotion, individual microzones exhibited large-scale, synchronized Ca2+ spiking. This is a mesoscopic neural dynamic missed by prior techniques for studying the brain at other length scales. Overall, the integrated microscope is a potentially transformative technology that permits distribution to many animals and enables diverse usages, such as portable diagnostics or microscope arrays for large-scale screens.