Microarray image analysis: background estimation using quantile and morphological filters

Background

Identification of protein interaction networks has received considerable attention in the post-genomic era. The currently available biochemical approaches used to detect protein-protein interactions are all time and labour intensive. Consequently there is a growing need for the development of computational tools that are capable of effectively identifying such interactions.

Results

Here we explain the development and implementation of a novel Protein-Protein Interaction Prediction Engine termed PIPE. This tool is capable of predicting protein-protein interactions for any target pair of the yeast Saccharomyces cerevisiae proteins from their primary structure and without the need for any additional information or predictions about the proteins. PIPE showed a sensitivity of 61% for detecting any yeast protein interaction with 89% specificity and an overall accuracy of 75%. This rate of success is comparable to those associated with the most commonly used biochemical techniques. Using PIPE, we identified a novel interaction between YGL227W (vid30) and YMR135C (gid8) yeast proteins. This lead us to the identification of a novel yeast complex that here we term vid30 complex (vid30c). The observed interaction was confirmed by tandem affinity purification (TAP tag), verifying the ability of PIPE to predict novel protein-protein interactions. We then used PIPE analysis to investigate the internal architecture of vid30c. It appeared from PIPE analysis that vid30c may consist of a core and a secondary component. Generation of yeast gene deletion strains combined with TAP tagging analysis indicated that the deletion of a member of the core component interfered with the formation of vid30c, however, deletion of a member of the secondary component had little effect (if any) on the formation of vid30c. Also, PIPE can be used to analyse yeast proteins for which TAP tagging fails, thereby allowing us to predict protein interactions that are not included in genome-wide yeast TAP tagging projects.

Conclusion

PIPE analysis can predict yeast protein-protein interactions. Also, PIPE analysis can be used to study the internal architecture of yeast protein complexes. The data also suggests that a finite set of short polypeptide signals seem to be responsible for the majority of the yeast protein-protein interactions.     

Comparative morphological studies of yeast cells using automated image analysis

The ultrastructural alterations in cells of Candida utilis caused by the influence of copper ions have been studied by means of quantitative image analysis. A model has been proposed which presents the following informations: The main effect of the copper ions is represented by an increase of the volume of the whole cell and of that part of the cell which consists of nucleus, vacuoles, and inclusions (particles and globules). Nevertheless, neither the absolute volume of mitochondria, nor the density of mitochondria are influenced by high concentrations of copper ions in the culture medium.     

Bayesian nonparametric quantile process regression and estimation of marginal quantile effects

Flexible estimation of multiple conditional quantiles is of interest in numerous applications, such as studying the effect of pregnancy-related factors on low and high birth weight. We propose a Bayesian nonparametric method to simultaneously estimate noncrossing, nonlinear quantile curves. We expand the conditional distribution function of the response in I-spline basis functions where the covariate-dependent coefficients are modeled using neural networks. By leveraging the approximation power of splines and neural networks, our model can approximate any continuous quantile function. Compared to existing models, our model estimates all rather than a finite subset of quantiles, scales well to high dimensions, and accounts for estimation uncertainty. While the model is arbitrarily flexible, interpretable marginal quantile effects are estimated using accumulative local effect plots and variable importance measures. A simulation study shows that our model can better recover quantiles of the response distribution when the data are sparse, and an analysis of birth weight data is presented.     

Microarray analysis of differentially expressed background genes in rats following hemorrhagic shock

To uncover the contribution of the diversity of the genetic backgrounds to the pathogenesis of hemorrhagic shock, we employed male Sprague-Dawley rats to establish a controlled 2.5 ml/100 g total body weight fixed-volume hemorrhagic shock and left lobular hepatectomy model. RNA was isolated from the liver samples taken from the rats (survival group: rats survived over 24 h after shock; and dead group: rats died within 1 h after shock, n = 3 per group), and subjected to microarray using the illumina^TM chips for rat cDNA (27,342 genes, >700,000 probes). The results demonstrated that the rats had about 50% survival rate and 100 genes were identified differentially expressed in the two groups. Of these genes, 47 genes were up-regulated and 53 genes down-regulated. Real-time PCR confirmed the differential expression for Aldh1a1, Aldh1a7, Aoc3, Cyp26al, Hdc and Ephx2 genes. Pathway analysis revealed that these genes are involved in circadian rhythm, beta-Alanine metabolism, histidine metabolism, biosynthesis of unsaturated fatty acids, glycine, serine and threonine metabolism, vitamin B6 metabolism, as well as arginine and proline metabolism. Therefore, our study provided a global molecular view on the contribution of genetic backgrounds to the response to hemorrhagic shock.     

Microarray analysis using bioinformatics analysis audit trails (BAATs)

Bioinformatics analysis plays an integrative role in genomics and functional genomics. The ability to conduct quality managed, hypothesis-driven bioinformatics analysis with the plethora of data available is mandatory. Biological interpretation of this data is dependent on versions of databases, programs and the parameters used. Thus, tracking and auditing the analyses process is important. This paper outlines what we term Bioinformatics Analysis Audit Trails (BAATs) and describes YABI, a bioinformatics environment that implements BAATs. YABI can incorporate most bioinformatics tools within the same environment, making it a valuable resource.     

An automatic objective estimation of vascularization of normal and tumor-invaded brain tissue using image analysis

A rapid method for the estimation of vascularization of normal and tumor-invaded brain tissue is described. The method involves immunostaining of cryostat sections using antibodies directed against a vascular protein (type IV collagen) and an objective measurement of these vascular elements using an automatic image processing system. Significant differences between tissues of normal brains and two tumors were shown. The potential of this method with regard to forming an index of malignancy is discussed.     

A novel method for the estimation of soybean chlorophyll content using a smartphone and image analysis

The development of smartphones, specifically their cameras, and imaging technologies has enabled their use as sensors/measurement tools. Here we aimed to evaluate the applicability of a fast and noninvasive method for the estimation of total chlorophyll (Chl), Chl a, Chl b, and carotenoids (Car) content of soybean plants using a smartphone camera. Single leaf disc images were obtained using a smartphone camera. Subsequently, for the same leaf discs, a Chl meter was used to obtain the relative index of Chl and the photosynthetic pigments were then determined using a classic method. The RGB, HSB and CIELab color models were extracted from the smartphone images and correlated to Chl values obtained using a Chl meter and by a standard laboratory protocol. The smartphone camera was sensitive enough to capture successfully a broad range of Chl and Car contents seen in soybean leaves. Although there was a variation between color models, some of the proposed regressions (e.g., the S and b index from HSB and Lab color models and NRI [RGB model]) were very close to the Chl meter values. Based on our findings, smartphones can be used for rapid and accurate estimation of soybean and Car contents in soybean leaves.     

A censored quantile regression approach for relative survival analysis: Relative survival quantile regression

We propose a censored quantile regression model for the analysis of relative survival data. We create a hybrid data set consisting of the study observations and counterpart randomly sampled pseudopopulation observations imputed from population life tables that adjust for expected mortality. We then fit a censored quantile regression model to the hybrid data incorporating demographic variables (e.g., age, biologic sex, calendar time) corresponding to the population life tables of demographically-similar individuals, a population versus study covariate, and its interactions with the variables of interest. These latter variables can be interpreted as relative survival parameters that depict the differences in failure quantiles between the study participants and their population counterparts.     

Analysis of morphological relationship between micro- and macromorphology of Mortierella species using a flow-through chamber coupled with image analysis

Using a flow-through chamber coupled with image analysis, the morphological parameters of 11 Mortierella species were quantified, and the relationship between micro- and macromorphology was investigated. On potato-dextrose-agar plates, 5 species formed rose petal-like colonies, 3 formed large round colonies, and 3 formed donut-like colonies. By observing micromorphology in a flow-through chamber, fungi were divided into 3 groups, classified according to morphological parameters: (i) a group with a high branch formation rate (q(b): tip/microm/h) and a low tip extension rate (q(tip): microm/tip/h); (ii) a group with a low branch formation rate and a high tip extension rate; and (iii) a group intermediate between the former and the latter groups. In suspension culture, group (i) fungi formed a hyphal bundle with a pulpy pellet-like morphology and a pellet core. In contrast, group (ii) fungi showed an aggregation of hyphae without the pellet core. In a narrow-specific hyphal growth rate (mu(l)) range (0.35-0.45 h(-1)), a higher branch formation rate led to increased hyphal branching, resulting in the formation of a hyphal bundle with a pulpy pellet-like morphology and a pellet core. When the branch formation rate was lower than 2 x 10(-3) tips/microm/h, the mycelia formed less branched but longer hypha. Our study surmises that a micromorphology consisting of a high hyphal growth rate (0.4 h(-1)), low tip extension rate (20 tips/microm/h), and high branch formation rate (8 x 10(-3) tips/microm/h) forms the suitable macromorphology for arachidonic acid production.     

Smooth quantile ratio estimation with regression: estimating medical expenditures for smoking-attributable diseases

The methodological development of this paper is motivated by a common problem in econometrics where we are interested in estimating the difference in the average expenditures between two populations, say with and without a disease, as a function of the covariates. For example, let Y(1) and Y(2) be two non-negative random variables denoting the health expenditures for cases and controls. Smooth Quantile Ratio Estimation (SQUARE) is a novel approach for estimating Delta=E[Y(1)] - E[Y(2)] by smoothing across percentiles the log-transformed ratio of the two quantile functions. Dominici et al. (2005) have shown that SQUARE defines a large class of estimators of Delta, is more efficient than common parametric and nonparametric estimators of Delta, and is consistent and asymptotically normal. However, in applications it is often desirable to estimate Delta(x)=E[Y(1)|x]--E[Y(2)|x], that is, the difference in means as a function of x. In this paper we extend SQUARE to a regression model and we introduce a two-part regression SQUARE for estimating Delta(x) as a function of x. We use the first part of the model to estimate the probability of incurring any costs and the second part of the model to estimate the mean difference in health expenditures, given that a nonzero cost is observed. In the second part of the model, we apply the basic definition of SQUARE for positive costs to compare expenditures for the cases and controls having 'similar' covariate profiles. We determine strata of cases and control with 'similar' covariate profiles by the use of propensity score matching. We then apply two-part regression SQUARE to the 1987 National Medicare Expenditure Survey to estimate the difference Delta(x) between persons suffering from smoking-attributable diseases and persons without these diseases as a function of the propensity of getting the disease. Using a simulation study, we compare frequentist properties of two-part regression SQUARE with maximum likelihood estimators for the log-transformed expenditures.     

Tag Frequency Difference: Rapid estimation of image set relevance for species occurrence data using general-purpose image classifiers

iEcology is used to supplement traditional ecological data by sourcing large quantities of media from the internet. Images and their metadata are widely available online and can provide information on species occurrence, behaviour and visible traits. However, this data is inherently noisy and data quality varies significantly between sources. Many iEcology studies utilise data from a single source for simplicity and efficiency. Hence, a tool to compare the suitability of different media sources in addressing a particular research question is needed.We provide a simple, novel way to estimate the fraction of images within multiple unverified datasets that potentially depict a specified target fauna. Our method, the Sum of Tag Frequency Differences (STFD), uses any pretrained, general-purpose image classifier. One of the method's innovations is that it does not require training the classifier to recognise the target fauna. Instead, STFD analyses the frequency of the generic text-tags returned by a classifier for multiple datasets and compares them to the corresponding frequencies of an authoritative image dataset that depicts only the target organism. From this comparison, STFD allows us to deduce the fraction of images of the target in unverified datasets.To validate the STFD approach, we processed images from five sources: Flickr, iNaturalist, Instagram, Reddit and Twitter. For each media source, we conducted an STFD analysis of three fauna invasive to Australia: Cane toads (Rhinella marina), German wasps (Vespula germanica), and the higher-level colloquial taxonomic classification, “wild rabbits”. We found the STFD provided an accurate assessment of image source relevance across all data sources and target organisms. This was demonstrated by the consistent, very strong correlation (toads r ≥0.97, wasps r ≥0.95, wild rabbits≥ 0.95) between STFD predictions, and the fraction of target images in a source dataset observed by a human expert.The STFD provides a low-cost, simple and accurate comparison of the relevance of online image sources to specific fauna for iEcology applications. It does not require expertise in machine learning or training neural-network species-specific classifiers. The method enables researchers to assess multiple image sources to select those warranting detailed investigation for the development of tools for web-scraping, citizen science campaigns, further monitoring or analysis.     

Cell identification and sizing using digital image analysis for estimation of cell biomass in High Rate Algal Ponds

Current environmental concerns make estimation of microbial biomass apriority for monitoring purposes and to advance scientific understanding. Thispaper considers problems associated with algal cell imaging and measurement forcell biomass estimation in samples from high rate algal ponds. In a complexsystem, the only way of measuring microbial activity is to measure theindividual cells and estimate biovolumes. Accurate biomass determinationsdemanddirect microscopic counting and measurement of the sizes of individualmicrobialcells taken from known volumes of water. The system used for routinemeasurementat the laboratory where the images were generated, based on standard microscopeequipment, is only suitable for treatment of well dispersed specimens.Differential interference contrast (DIC) microscopy, on the other hand, offersthe best solution for optical enhancement of cell contrast, and produces animage with well defined edges, yet presents a great challenge to routine cellidentification by digital image analysis, owing to the bas-relief type imageproduced. The paper outlines several image analysis methods developedspecifically for this purpose, and presents illustrative results.     

The use of image analysis for morphological measurements on filamentous microorganisms

Characterization of mycelial morphology is important for the design and operation of filamentous fermentations. Initial investigations have been made of semiautomated image analysis as a replacement for a digitizing-table method. It was shown that the image analysis method was more precise than the digitizing-table method used, although this extra precision is unimportant in this application. On average, image analysis gave mean hyphal lengths 6% greater than digitizing, because the latter used chord lengths to represent are lengths. For short branches image analysis was less accurate. In time and convenience the image analysis method had an overwhelming advantage, and this advantage might be enhanced by full automation. The resulting ability to characterize mycelial morphology rapidly would permit such characterization to be used routinely in studies of filamentous fermentations.