In Situ Hybridization Images Research Articles

Review of In Situ Hybridization (ISH) Stain Images Using Computational Techniques.

Recent advancements in medical imaging have greatly enhanced the application of computational techniques in digital pathology, particularly for the classification of breast cancer using in situ hybridization (ISH) imaging. HER2 amplification, a key prognostic marker in 20-25% of breast cancers, can be assessed through alterations in gene copy number or protein expression. However, challenges persist due to the heterogeneity of nuclear regions and complexities in cancer biomarker detection. This review examines semi-automated and fully automated computational methods for analyzing ISH images with a focus on HER2 gene amplification. Literature from 1997 to 2023 is analyzed, emphasizing silver-enhanced in situ hybridization (SISH) and its integration with image processing and machine learning techniques. Both conventional machine learning approaches and recent advances in deep learning are compared. The review reveals that automated ISH analysis in combination with bright-field microscopy provides a cost-effective and scalable solution for routine pathology. The integration of deep learning techniques shows promise in improving accuracy over conventional methods, although there are limitations related to data variability and computational demands. Automated ISH analysis can reduce manual labor and increase diagnostic accuracy. Future research should focus on refining these computational methods, particularly in handling the complex nature of HER2 status evaluation, and integrate best practices to further enhance clinical adoption of these techniques.

HER2 in situ hybridization testing in breast cancer: Applying algorithm-assisted assessment to reduce interobserver variability in difficult cases.

e13561 Background: Accurate assessment of HER2 expression by HER2 immunohistochemistry and in-situ hybridization (ISH) is critical for the management of patients with breast cancer. The revised 2018 ASCO/CAP guidelines defined 5 subgroups based on HER2 expression and copy number. Few reports have investigated interobserver variability when applying the revised guidelines. Digital algorithms may have a role in assisting pathologists in assessing cases with equivocal and uncommon patterns of HER2 amplification. This study was designed to evaluate interobserver variability in the gold standard manual assessment of HER2 ISH, particularly in less common HER2 ISH groups. Subsequently, we evaluated whether a new digital image analysis algorithm (Roche uPath HER2 Dual ISH image analysis) could reduce interobserver variability and potentially improve patient selection for targeted anti-HER2 therapy. Methods: Of 2982 breast cancer cases, we derived an enriched cohort of 99 patient cases with HER2 IHC 2+ or 3+ and selective ISH results (groups 2-4, evidence of HER2 heterogeneity, borderline negative group 5 and borderline positive group 1). As per guidelines, a minimum of 20 tumour cells per slide were manually and independently counted by 4 anatomical pathologists. After a washout period and second round of de-identification, the slides were scanned (DP200; Roche) and re-scored using the digital algorithm. Results: Pathologist reporting of patient HER2 status (HER2 positive vs. HER2 negative) using manual microscopy demonstrated significant interobserver variability, with a Fleiss’s kappa value of 0.471 (95% CI: 0.378-0.564; P= 1.24E-21) indicating fair-moderate agreement. Interobserver variability in manual scoring of HER2 ISH at the level of HER2 group designation (groups 1-5) showed poor-moderate reliability between pathologists, with an intraclass correlation coefficient (ICC) of 0.526 (95% CI: 0.407-0.641; P= 9.28E-21). Repeat analysis of the cohort using algorithm-assisted quantification resulted in moderate-substantial concordance in HER2 patient status reporting, with a Fleiss’s kappa value of 0.666 (95% CI: 0.572-0.759; P= 1.70E-40). Algorithm-assisted quantification resulted in moderate-good agreement of HER2 group designation, with an ICC of 0.763 (95% CI: 0.683-0.832; P= 8.76E-48). In sub-group analysis, use of the algorithm improved concordance particularly in groups 2,4 and 5. Time to report cases was also significantly reduced with the use of the algorithm. Conclusions: Our study demonstrates the potential to improve the concordance of HER2 amplification status reporting in less common HER2 subgroups using a digital algorithm and potentially improve therapy selection and outcomes for patients with HER2-low and borderline HER2-amplified breast cancers.

Deep Low-Shot Learning for Biological Image Classification and Visualization From Limited Training Samples.

Predictive modeling is useful but very challenging in biological image analysis due to the high cost of obtaining and labeling training data. For example, in the study of gene interaction and regulation in Drosophila embryogenesis, the analysis is most biologically meaningful when in situ hybridization (ISH) gene expression pattern images from the same developmental stage are compared. However, labeling training data with precise stages is very time-consuming even for developmental biologists. Thus, a critical challenge is how to build accurate computational models for precise developmental stage classification from limited training samples. In addition, identification and visualization of developmental landmarks are required to enable biologists to interpret prediction results and calibrate models. To address these challenges, we propose a deep two-step low-shot learning framework to accurately classify ISH images using limited training images. Specifically, to enable accurate model training on limited training samples, we formulate the task as a deep low-shot learning problem and develop a novel two-step learning approach, including data-level learning and feature-level learning. We use a deep residual network as our base model and achieve improved performance in the precise stage prediction task of ISH images. Furthermore, the deep model can be interpreted by computing saliency maps, which consists of pixel-wise contributions of an image to its prediction result. In our task, saliency maps are used to assist the identification and visualization of developmental landmarks. Our experimental results show that the proposed model can not only make accurate predictions but also yield biologically meaningful interpretations. We anticipate our methods to be easily generalizable to other biological image classification tasks with small training datasets. Our open-source code is available at https://github.com/divelab/lsl-fly.

Cell-type-specific densities in mouse somatosensory cortex derived from scRNA-seq and in situ RNA hybridization.

Cells in the mammalian cerebral cortex exhibit layer-dependent patterns in their distribution. Classical methods of determining cell type distributions typically employ a painstaking process of large-scale sampling and characterization of cellular composition. We found that by combining in situ hybridization (ISH) images with cell-type-specific transcriptomes, position-dependent cortical composition in P56 mouse could be estimated in the somatosensory cortex. The method uses ISH images from the Allen Institute for Brain Science. There are two novel aspects of the methodology. First, it is not necessary to select a subset of genes that are particular for a cell type of interest, nor is it necessary to only use ISH images with low variability among samples. Second, the method also compensated for differences in soma size and incompleteness of the transcriptomes. The soma size compensation is particularly important in order to obtain quantitative estimates since relying on bulk expression alone would overestimate the contribution of larger cells. Predicted distributions of broader classes of cell types agreed with literature distributions. The primary result is that there is a high degree of substructure in the distribution of transcriptomic types beyond the resolution of layers. Furthermore, transcriptomic cell types each exhibited characteristic soma size distributions. Results suggest that the method could also be employed to assign transcriptomic cell types to well-aligned image sets in the entire brain.

Endoplasmic Reticulum Stress in Colonic Mucosa of Ulcerative Colitis Patients Is Mediated by PERK and IRE1 Pathway Activation.

Ulcerative colitis (UC) is characterized by a chronic overproduction of proinflammatory cytokines. During an acute phase, the endoplasmic reticulum (ER) is overloaded and the protein folding process is impaired, a condition named ER stress. This state induces a response (unfolded protein response (UPR)), initiated by the activation of IRE1/Xbp-1, PERK/eIF2α, and ATF6 pathways, which has previously been linked to intestinal inflammation in experimental models. ER stress and UPR activation trigger the activation of proinflammatory, autophagy, and apoptosis genes, in addition to promoting protein degradation. Therefore, the goal of this study was to evaluate the activation of ER stress and UPR in colonic mucosa of UC patients. Patient and Methods. Transcriptional analysis of ER stress- and UPR-related genes was performed by qPCR from intestinal mucosa of patients with UC. We also performed in situ hybridization (ISH) and immunohistochemistry (IHQ) of PERK/eIF2α and IRE1/Xbp-1 pathways and UPR-related chaperones. Results. We first evaluated inflammatory genes via qPCR, and we observed that all analyzed proinflammatory transcripts were upregulated in UC patients. ISH and IHQ images showed that ER stress is activated via PERK/eIF2α and IRE1/Xbp-1 pathways not only in intestinal epithelial cells but also in cells of the lamina propria of UC colonic mucosa. Transcriptional analysis confirmed that EIF2AK3 was upregulated in UC patients. UPR-related genes, such as ATF3, STC2, and DDIT3, along with the chaperones and cochaperones DNAJC3, CALR, HSP90B1, and HSPA5, were also upregulated in UC patients. In addition, we observed that proapoptotic and autophagy genes (Bax and ATG6L1, respectively) were also upregulated. Conclusion. Our results suggest that ER stress and UPR are indeed activated in UC patients and this may contribute to the chronic inflammatory process seen in UC. The increased apoptosis and autophagy markers further support the activation of these findings once they are activated to counterbalance tissue damage. These findings provide new insights into the molecular mechanisms that maintain UC activity and open new possibilities to attenuate intestinal inflammation.

Evaluation of deep convolutional neural networks for in situ hybridization gene expression image representation.

High resolution in situ hybridization (ISH) images of the brain capture spatial gene expression at cellular resolution. These spatial profiles are key to understanding brain organization at the molecular level. Previously, manual qualitative scoring and informatics pipelines have been applied to ISH images to determine expression intensity and pattern. To better capture the complex patterns of gene expression in the human cerebral cortex, we applied a machine learning approach. We propose gene re-identification as a contrastive learning task to compute representations of ISH images. We train our model on an ISH dataset of ~1,000 genes obtained from postmortem samples from 42 individuals. This model reaches a gene re-identification rate of 38.3%, a 13x improvement over random chance. We find that the learned embeddings predict expression intensity and pattern. To test generalization, we generated embeddings in a second dataset that assayed the expression of 78 genes in 53 individuals. In this set of images, 60.2% of genes are re-identified, suggesting the model is robust. Importantly, this dataset assayed expression in individuals diagnosed with schizophrenia. Gene and donor-specific embeddings from the model predict schizophrenia diagnosis at levels similar to that reached with demographic information. Mutations in the most discriminative gene, Sodium Voltage-Gated Channel Beta Subunit 4 (SCN4B), may help understand cardiovascular associations with schizophrenia and its treatment. We have publicly released our source code, embeddings, and models to spur further application to spatial transcriptomics. In summary, we propose and evaluate gene re-identification as a machine learning task to represent ISH gene expression images.

Computer-aided scoring of erb-b2 receptor tyrosine kinase 2 (HER2) gene amplification status in breast cancer

BackgroundIdentification of HER2 protein overexpression and/or amplification of the HER2 gene are required to qualify breast cancer patients for HER2 targeted therapies. In situ hybridization (ISH) assays that identify HER2 gene amplification function as a stand-alone test for determination of HER2 status and rely on the manual quantification of the number of HER2 genes and copies of chromosome 17 to determine HER2 amplification. MethodsTo assist pathologists, we have developed the uPath HER2 Dual ISH Image Analysis for Breast (uPath HER2 DISH IA) algorithm, as an adjunctive aid in the determination of HER2 gene status in breast cancer specimens. The objective of this study was to compare uPath HER2 DISH image analysis vs manual read scoring of VENTANA HER2 DISH-stained breast carcinoma specimens with ground truth (GT) gene status as the reference. Three reader pathologists reviewed 220, formalin-fixed, paraffin-embedded (FFPE) breast cancer cases by both manual and uPath HER2 DISH IA methods. Scoring results from manual read (MR) and computer-assisted scores (image analysis, IA) were compared against the GT gene status generated by consensus of a panel of pathologists. The differences in agreement rates of HER2 gene status between manual, computer-assisted, and GT gene status were determined. ResultsThe positive percent agreement (PPA) and negative percent agreement (NPA) rates for image analysis (IA) vs GT were 97.2% (95% confidence interval [CI]: 95.0, 99.3) and 94.3% (95% CI: 90.8, 97.3) respectively. Comparison of agreement rates showed that the lower bounds of the 95% CIs for the difference of PPA and NPA for IA vs MR were –0.9% and –6.2%, respectively. Further, inter- and intra-reader agreement rates in the IA method were observed with point estimates of at least 96.7%. ConclusionsOverall, our data show that the uPath HER2 DISH IA is non-inferior to manual scoring and supports its use as an aid for pathologists in routine diagnosis of breast cancer.

Deep convolutional neural networks for annotating gene expression patterns in the mouse brain.

BackgroundProfiling gene expression in brain structures at various spatial and temporal scales is essential to understanding how genes regulate the development of brain structures. The Allen Developing Mouse Brain Atlas provides high-resolution 3-D in situ hybridization (ISH) gene expression patterns in multiple developing stages of the mouse brain. Currently, the ISH images are annotated with anatomical terms manually. In this paper, we propose a computational approach to annotate gene expression pattern images in the mouse brain at various structural levels over the course of development.ResultsWe applied deep convolutional neural network that was trained on a large set of natural images to extract features from the ISH images of developing mouse brain. As a baseline representation, we applied invariant image feature descriptors to capture local statistics from ISH images and used the bag-of-words approach to build image-level representations. Both types of features from multiple ISH image sections of the entire brain were then combined to build 3-D, brain-wide gene expression representations. We employed regularized learning methods for discriminating gene expression patterns in different brain structures. Results show that our approach of using convolutional model as feature extractors achieved superior performance in annotating gene expression patterns at multiple levels of brain structures throughout four developing ages. Overall, we achieved average AUC of 0.894 ± 0.014, as compared with 0.820 ± 0.046 yielded by the bag-of-words approach.ConclusionsDeep convolutional neural network model trained on natural image sets and applied to gene expression pattern annotation tasks yielded superior performance, demonstrating its transfer learning property is applicable to such biological image sets.

Automated identification of cell-type-specific genes in the mouse brain by image computing of expression patterns.

BackgroundDifferential gene expression patterns in cells of the mammalian brain result in the morphological, connectional, and functional diversity of cells. A wide variety of studies have shown that certain genes are expressed only in specific cell-types. Analysis of cell-type-specific gene expression patterns can provide insights into the relationship between genes, connectivity, brain regions, and cell-types. However, automated methods for identifying cell-type-specific genes are lacking to date.ResultsHere, we describe a set of computational methods for identifying cell-type-specific genes in the mouse brain by automated image computing of in situ hybridization (ISH) expression patterns. We applied invariant image feature descriptors to capture local gene expression information from cellular-resolution ISH images. We then built image-level representations by applying vector quantization on the image descriptors. We employed regularized learning methods for classifying genes specifically expressed in different brain cell-types. These methods can also rank image features based on their discriminative power. We used a data set of 2,872 genes from the Allen Brain Atlas in the experiments. Results showed that our methods are predictive of cell-type-specificity of genes. Our classifiers achieved AUC values of approximately 87% when the enrichment level is set to 20. In addition, we showed that the highly-ranked image features captured the relationship between cell-types.ConclusionsOverall, our results showed that automated image computing methods could potentially be used to identify cell-type-specific genes in the mouse brain.

Background coloration of squamous epithelium in esophago-pharyngeal squamous cell carcinoma: what causes the color change?

ObjectivesThis study aims to clarify the cause of background coloration in the epithelia between each dilated intra papillary capillary loop in esophago-pharyngeal squamous cell carcinoma.DesignThis is a single center retrospective study including 124 patients with 160 lesions who underwent esophagogastroduodenoscopy in Nagasaki University Hospital from September 2007 to March 2012; a detailed comparison between endoscopic images and pathology was performed. Immunohistological assessment using anti-human hemoglobin antibody (anti-Hb Ab) was performed to verify the presence of hemoglobin (Hb) component in the cancer cells. Real-time polymerase chain reaction (RT-PCR) and in situ hybridization (ISH) on Hb-β mRNA were performed to assess the production of Hb component within the cancer cells.ResultsA strong positivity for anti-Hb Ab was observed in the squamous cell carcinoma area, whereas non-cancerous mucosa showed no immunopositivity for Hb. The concordance rate between anti-Hb Ab immunoreactivity and the presence of BC was as high as 80.9%. The amount of Hb-β mRNA expression was three times higher in cancer tissues compared with the surrounding non-cancerous mucosa. ISH images showed that the expression exclusively occurred in cancer cells, indicating that Hb is probably produced within cancer cells.ConclusionsThe background coloration observed is partly due to an extravascular component of Hb. RT-PCR and ISH analyses indicate that Hb is produced within cancer cells.

GINI: From ISH Images to Gene Interaction Networks

Accurate inference of molecular and functional interactions among genes, especially in multicellular organisms such as Drosophila, often requires statistical analysis of correlations not only between the magnitudes of gene expressions, but also between their temporal-spatial patterns. The ISH (in-situ-hybridization)-based gene expression micro-imaging technology offers an effective approach to perform large-scale spatial-temporal profiling of whole-body mRNA abundance. However, analytical tools for discovering gene interactions from such data remain an open challenge due to various reasons, including difficulties in extracting canonical representations of gene activities from images, and in inference of statistically meaningful networks from such representations. In this paper, we present GINI, a machine learning system for inferring gene interaction networks from Drosophila embryonic ISH images. GINI builds on a computer-vision-inspired vector-space representation of the spatial pattern of gene expression in ISH images, enabled by our recently developed system; and a new multi-instance-kernel algorithm that learns a sparse Markov network model, in which, every gene (i.e., node) in the network is represented by a vector-valued spatial pattern rather than a scalar-valued gene intensity as in conventional approaches such as a Gaussian graphical model. By capturing the notion of spatial similarity of gene expression, and at the same time properly taking into account the presence of multiple images per gene via multi-instance kernels, GINI is well-positioned to infer statistically sound, and biologically meaningful gene interaction networks from image data. Using both synthetic data and a small manually curated data set, we demonstrate the effectiveness of our approach in network building. Furthermore, we report results on a large publicly available collection of Drosophila embryonic ISH images from the Berkeley Drosophila Genome Project, where GINI makes novel and interesting predictions of gene interactions. Software for GINI is available at http://sailing.cs.cmu.edu/Drosophila_ISH_images/

FuncISH: learning a functional representation of neural ISH images

Motivation: High-spatial resolution imaging datasets of mammalian brains have recently become available in unprecedented amounts. Images now reveal highly complex patterns of gene expression varying on multiple scales. The challenge in analyzing these images is both in extracting the patterns that are most relevant functionally and in providing a meaningful representation that allows neuroscientists to interpret the extracted patterns.Results: Here, we present FuncISH—a method to learn functional representations of neural in situ hybridization (ISH) images. We represent images using a histogram of local descriptors in several scales, and we use this representation to learn detectors of functional (GO) categories for every image. As a result, each image is represented as a point in a low-dimensional space whose axes correspond to meaningful functional annotations. The resulting representations define similarities between ISH images that can be easily explained by functional categories. We applied our method to the genomic set of mouse neural ISH images available at the Allen Brain Atlas, finding that most neural biological processes can be inferred from spatial expression patterns with high accuracy. Using functional representations, we predict several gene interaction properties, such as protein–protein interactions and cell-type specificity, more accurately than competing methods based on global correlations. We used FuncISH to identify similar expression patterns of GABAergic neuronal markers that were not previously identified and to infer new gene function based on image–image similarities.Contact: noalis@gmail.comSupplementary information: Supplementary data are available at Bioinformatics online.

Altered gene expression in the dorsolateral prefrontal cortex of individuals with schizophrenia

The underlying pathology of schizophrenia (SZ) is likely as heterogeneous as its symptomatology. A variety of cortical and subcortical regions, including the prefrontal cortex, have been implicated in its pathology, and a number of genes have been identified as risk factors for disease development. We used in situ hybridization (ISH) to examine the expression of 58 genes in the dorsolateral prefrontal cortex (DLPFC, comprised of Brodmann areas 9 and 46) from 19 individuals with a premorbid diagnosis of SZ and 33 control individuals. Genes were selected based on: (1) previous identification as risk factors for SZ; (2) cell type markers or (3) laminar markers. Cell density and staining intensity were compared in the DLPFC, as well as separately in Brodmann areas 9 and 46. The expression patterns of a variety of genes, many of which are associated with the GABAergic system, were altered in SZ when compared with controls. Additional genes, including C8orf79 and NR4A2, showed alterations in cell density or staining intensity between the groups, highlighting the need for additional studies. Alterations were, with only a few exceptions, limited to Brodmann area 9, suggesting regional specificity of pathology in the DLPFC. Our results agree with previous studies on the GABAergic involvement in SZ, and suggest that areas 9 and 46 may be differentially affected in the disease. This study also highlights additional genes that may be altered in SZ, and indicates that these potentially interesting genes can be identified by ISH and high-throughput image analysis techniques.

Localizing Genes to Cerebellar Layers by Classifying ISH Images

Gene expression controls how the brain develops and functions. Understanding control processes in the brain is particularly hard since they involve numerous types of neurons and glia, and very little is known about which genes are expressed in which cells and brain layers. Here we describe an approach to detect genes whose expression is primarily localized to a specific brain layer and apply it to the mouse cerebellum. We learn typical spatial patterns of expression from a few markers that are known to be localized to specific layers, and use these patterns to predict localization for new genes. We analyze images of in-situ hybridization (ISH) experiments, which we represent using histograms of local binary patterns (LBP) and train image classifiers and gene classifiers for four layers of the cerebellum: the Purkinje, granular, molecular and white matter layer. On held-out data, the layer classifiers achieve accuracy above 94% (AUC) by representing each image at multiple scales and by combining multiple image scores into a single gene-level decision. When applied to the full mouse genome, the classifiers predict specific layer localization for hundreds of new genes in the Purkinje and granular layers. Many genes localized to the Purkinje layer are likely to be expressed in astrocytes, and many others are involved in lipid metabolism, possibly due to the unusual size of Purkinje cells.

Allen Brain Atlas: an integrated spatio-temporal portal for exploring the central nervous system

The Allen Brain Atlas (http://www.brain-map.org) provides a unique online public resource integrating extensive gene expression data, connectivity data and neuroanatomical information with powerful search and viewing tools for the adult and developing brain in mouse, human and non-human primate. Here, we review the resources available at the Allen Brain Atlas, describing each product and data type [such as in situ hybridization (ISH) and supporting histology, microarray, RNA sequencing, reference atlases, projection mapping and magnetic resonance imaging]. In addition, standardized and unique features in the web applications are described that enable users to search and mine the various data sets. Features include both simple and sophisticated methods for gene searches, colorimetric and fluorescent ISH image viewers, graphical displays of ISH, microarray and RNA sequencing data, Brain Explorer software for 3D navigation of anatomy and gene expression, and an interactive reference atlas viewer. In addition, cross data set searches enable users to query multiple Allen Brain Atlas data sets simultaneously. All of the Allen Brain Atlas resources can be accessed through the Allen Brain Atlas data portal.

SPEX2: automated concise extraction of spatial gene expression patterns from Fly embryo ISH images

Motivation: Microarray profiling of mRNA abundance is often ill suited for temporal–spatial analysis of gene expressions in multicellular organisms such as Drosophila. Recent progress in image-based genome-scale profiling of whole-body mRNA patterns via in situ hybridization (ISH) calls for development of accurate and automatic image analysis systems to facilitate efficient mining of complex temporal–spatial mRNA patterns, which will be essential for functional genomics and network inference in higher organisms.Results: We present SPEX2, an automatic system for embryonic ISH image processing, which can extract, transform, compare, classify and cluster spatial gene expression patterns in Drosophila embryos. Our pipeline for gene expression pattern extraction outputs the precise spatial locations and strengths of the gene expression. We performed experiments on the largest publicly available collection of Drosophila ISH images, and show that our method achieves excellent performance in automatic image annotation, and also finds clusters that are significantly enriched, both for gene ontology functional annotations, and for annotation terms from a controlled vocabulary used by human curators to describe these images.Availability: Software will be available at http://www.sailing.cs.cmu.edu/Contact: epxing@cs.cmu.eduSupplementary information: Supplementary data are avilable at Bioinformatics online.

Comparative Analysis of Selenocysteine Machinery and Selenoproteome Gene Expression in Mouse Brain Identifies Neurons as Key Functional Sites of Selenium in Mammals

Although dietary selenium (Se) deficiency results in phenotypes associated with selenoprotein depletion in various organs, the brain is protected from Se loss. To address the basis for the critical role of Se in brain function, we carried out comparative gene expression analyses for the complete selenoproteome and associated biosynthetic factors. Using the Allen Brain Atlas, we evaluated 159 regions of adult mouse brain and provided experimental analyses of selected selenoproteins. All 24 selenoprotein mRNAs were expressed in the mouse brain. Most strikingly, neurons in olfactory bulb, hippocampus, cerebral cortex, and cerebellar cortex were exceptionally rich in selenoprotein gene expression, in particular in GPx4, SelK, SelM, SelW, and Sep15. Over half of the selenoprotein genes were also expressed in the choroid plexus. A unique expression pattern was observed for one of the highly expressed selenoprotein genes, SelP, which we suggest to provide neurons with Se. Cluster analysis of the expression data linked certain selenoproteins and selenocysteine machinery genes and suggested functional linkages among selenoproteins, such as that between SelM and Sep15. Overall, this study suggests that the main functions of selenium in mammals are confined to certain neurons in the brain.

Automatic image analysis for gene expression patterns of fly embryos

BackgroundStaining the mRNA of a gene via in situ hybridization (ISH) during the development of a D. melanogaster embryo delivers the detailed spatio-temporal pattern of expression of the gene. Many biological problems such as the detection of co-expressed genes, co-regulated genes, and transcription factor binding motifs rely heavily on the analyses of these image patterns. The increasing availability of ISH image data motivates the development of automated computational approaches to the analysis of gene expression patterns.ResultsWe have developed algorithms and associated software that extracts a feature representation of a gene expression pattern from an ISH image, that clusters genes sharing the same spatio-temporal pattern of expression, that suggests transcription factor binding (TFB) site motifs for genes that appear to be co-regulated (based on the clustering), and that automatically identifies the anatomical regions that express a gene given a training set of annotations. In fact, we developed three different feature representations, based on Gaussian Mixture Models (GMM), Principal Component Analysis (PCA), and wavelet functions, each having different merits with respect to the tasks above. For clustering image patterns, we developed a minimum spanning tree method (MSTCUT), and for proposing TFB sites we used standard motif finders on clustered/co-expressed genes with the added twist of requiring conservation across the genomes of 8 related fly species. Lastly, we trained a suite of binary-classifiers, one for each anatomical annotation term in a controlled vocabulary or ontology that operate on the wavelet feature representation. We report the results of applying these methods to the Berkeley Drosophila Genome Project (BDGP) gene expression database.ConclusionOur automatic image analysis methods recapitulate known co-regulated genes and give correct developmental-stage classifications with 99+% accuracy, despite variations in morphology, orientation, and focal plane suggesting that these techniques form a set of useful tools for the large-scale computational analysis of fly embryonic gene expression patterns.

Semiautomated Multiplexed Quantum Dot-Based in Situ Hybridization and Spectral Deconvolution

Gene expression profiling has identified several potentially useful gene signatures for predicting outcome or for selecting targeted therapy. However, these signatures have been developed in fresh or frozen tissue, and there is a need to apply them to routinely processed samples. Here, we demonstrate the feasibility of a potentially high-throughput methodology combining automated in situ hybridization with quantum dot-labeled oligonucleotide probes followed by spectral imaging for the detection and subsequent deconvolution of multiple signals. This method is semiautomated and quantitative and can be applied to formalin-fixed, paraffin-embedded tissues. We have combined dual in situ hybridization with immunohistochemistry, enabling simultaneous measurement of gene expression and cell lineage determination. The technique achieves levels of sensitivity and specificity sufficient for the potential application of known expression signatures to biopsy specimens in a semiquantitative way, and the semiautomated nature of the method enables application to high-throughput studies.

GEISHA, a whole‐mount in situ hybridization gene expression screen in chicken embryos

Despite the increasing quality and quantity of genomic sequence that is available to researchers, predicting gene function from sequence information remains a challenge. One method for obtaining rapid insight into potential functional roles of novel genes is through gene expression mapping. We have performed a high throughput whole-mount in situ hybridization (ISH) screen with chick embryos to identify novel, differentially expressed genes. Approximately 1,200 5' expressed sequence tags (ESTs) were generated from cDNA clones of a Hamburger and Hamilton (HH) stage 4-7 (late gastrula) chick embryo endoderm-mesoderm library. After screening to remove ubiquitously expressed cDNAs and internal clustering and after comparison to GenBank sequences, remaining cDNAs (representing both characterized and uncharacterized genes) were screened for expression in HH stage 3-14 embryos by automated high throughput ISH. Of 786 cDNAs for which ISH was successfully performed, approximately 30% showed ubiquitous expression, 40% were negative, and approximately 30% showed a restricted expression pattern. cDNAs were identified that showed restricted expression in every embryonic region, including the primitive streak, somites, developing cardiovascular system and neural tube/neural crest. A relational database was developed to hold all EST sequences, ISH images, and corresponding BLAST report information, and to enable browsing and querying of data. A user interface is freely accessible at http://geisha.biosci.arizona.edu. Results show that high throughput whole-mount ISH provides an effective approach for identifying novel genes that are differentially expressed in the developing chicken embryo.