Berkeley Drosophila Genome Project Research Articles

Unsupervised Heterogeneous Transfer Learning for Partial Co-occurrence Data

Heterogeneous transfer clustering contributes to improve the performance of target domain by using the co-occurrence data from different domains without any supervision. Existing works usually use a large of complete co-occurrence data to learn the projection functions mapping heterogeneous data to a common latent feature subspace. Given that in the real-world problems, complete and abundant co-occurrence data in the form of homogeneous transfer learning between the soured domain and target domain are hard to achieve, a heterogeneous transfer clustering method for partial co-occurrence data (HTCPC) is proposed here, to perform unsupervised learning to map the partial information obtained from the source domain onto objects in the target domain. Furthermore, to maximize the useful information to improve the clustering performance in target domain, the proposed HTCPC uses the deep matrix decomposition framework to maintain the multi-layer hidden feature representation and retain the complexity of the data hierarchy by adding the approximate orthogonal constraints, which can effectively strengthen the independence and minimal redundancy. From a series of experiments conducted on four datasets [Berkeley Drosophila Genome Project (BDGP), Devanagari Handwritten Character (DHC), Columbia University Image Library (COIL), and Notting-Hill (NH)], the results show that HTCPC outperforms the peers in the following aspects: our method constructs the hierarchical structure in the multi-layer latent representations and the proposed algorithm can reduce the redundancy and extract more useful knowledge for target domain.

In silico Characterization of a Candidate Protein from Aphid Gelling Saliva with Potential for Aphid Control in Plants.

Sheath or gelling saliva, secreted during feeding by aphids, is a hard material that supports the piercing mouthparts and remains in the plant after feeding. Solidification or gelling of the saliva might be due to the composition of amino acids in the constituent proteins, many of which probably interact with plant defenses. The complete complement of proteins in the gelling saliva are still unknown, although one sheath protein (SHP) has previously been identified as a potential candidate protein to control aphid feeding, but its structure and its physiochemical role remains obscure. The current study provides structural information and biochemical properties of the aphid sheath protein. The Sheath protein encoding gene was amplified from cDNA of the pea aphid (Acyrthosiphon pisum) through PCR using specific gene primers. Sequence was in silico characterized by using EXPASY, Berkeley Drosophila Genome Project (BDGP) Neural Network Promoter Prediction, BioEdit, Mega7, ProtParam, Phyre server, 3D LigandSite SMART, MEME and GSDS programs, available online. BLASTp analysis revealed that the sequenced gene was identical (100%) to the sequence from Acyrthosiphon pisum, with 87% identity to Metpolophium dirhodum and 84% identity to Sitobion avenae. Phylogenetically monocot feeders such as M. dirhodum and S. avenae are in a sister taxa to dicot feeders. In silico analysis of the sequence revealed that sheath protein has a molecular weight of 144 kDa and 50% of the protein is composed of only six amino acids, i.e., threonine, serine, aspartic acid, glutamic acid, isoleucine and tyrosine. The computed IP value revealed that sheath protein is acidic in nature. Ligand binding sites for sheath protein were predicted on residues 1123 and 1125 (isoleucine and glutamine, respectively). Metallic heterogens are also present in sheath protein that are iron, zinc and magnesium, respectively. It is conceivable that variation in the salivary gene sequences may reveal important biological information of relevance to the insect-plant interaction. Further exploration of insect salivary proteins, their composition and structure will provide powerful information, especially when these proteins are interacting with plant proteins, and specific information about the sheath protein, which is interacting with plants at a molecular/cellular level, will be important to progress strategies aimed specifically against sucking pests such as aphids.

AnnoFly: annotating Drosophila embryonic images based on an attention-enhanced RNN model.

In the post-genomic era, image-based transcriptomics have received huge attention, because the visualization of gene expression distribution is able to reveal spatial and temporal expression pattern, which is significantly important for understanding biological mechanisms. The Berkeley Drosophila Genome Project has collected a large-scale spatial gene expression database for studying Drosophila embryogenesis. Given the expression images, how to annotate them for the study of Drosophila embryonic development is the next urgent task. In order to speed up the labor-intensive labeling work, automatic tools are highly desired. However, conventional image annotation tools are not applicable here, because the labeling is at the gene-level rather than the image-level, where each gene is represented by a bag of multiple related images, showing a multi-instance phenomenon, and the image quality varies by image orientations and experiment batches. Moreover, different local regions of an image correspond to different CV annotation terms, i.e. an image has multiple labels. Designing an accurate annotation tool in such a multi-instance multi-label scenario is a very challenging task. To address these challenges, we develop a new annotator for the fruit fly embryonic images, called AnnoFly. Driven by an attention-enhanced RNN model, it can weight images of different qualities, so as to focus on the most informative image patterns. We assess the new model on three standard datasets. The experimental results reveal that the attention-based model provides a transparent approach for identifying the important images for labeling, and it substantially enhances the accuracy compared with the existing annotation methods, including both single-instance and multi-instance learning methods. http://www.csbio.sjtu.edu.cn/bioinf/annofly/. Supplementary data are available at Bioinformatics online.

FlyBase: introduction of the Drosophila melanogaster Release 6 reference genome assembly and large-scale migration of genome annotations.

Release 6, the latest reference genome assembly of the fruit fly Drosophila melanogaster, was released by the Berkeley Drosophila Genome Project in 2014; it replaces their previous Release 5 genome assembly, which had been the reference genome assembly for over 7 years. With the enormous amount of information now attached to the D. melanogaster genome in public repositories and individual laboratories, the replacement of the previous assembly by the new one is a major event requiring careful migration of annotations and genome-anchored data to the new, improved assembly. In this report, we describe the attributes of the new Release 6 reference genome assembly, the migration of FlyBase genome annotations to this new assembly, how genome features on this new assembly can be viewed in FlyBase (http://flybase.org) and how users can convert coordinates for their own data to the corresponding Release 6 coordinates.

Structural and functional characterization of the actin-1 gene promoter from the Antheraea pernyi (Lepidoptera: Saturniidae).

The Chinese oak silkworm, Antheraea pernyi , is an economically important insect of the Saturniidae family. In this study, genome walking was performed to obtain an A. pernyi actin promoter, which can be employed in transgenic or stable cell line expression systems. The putative promoter was analyzed by the online promoter analysis programs at the Berkeley Drosophila Genome Project and the Web Promoter Scan Service, which led to the recognition of several functional elements. With respect to these elements, a series of actin A1 promoter fragments with 5′-deletions were generated that were then used to construct different vectors expressing Green Fluorescent Protein (GFP). The plasmids were transfected into Sf9 cells and GFP expression was determined by observing GFP fluorescence in cells and by measuring GFP mRNA levels with real-time polymerase chain reaction. Sequence comparisons indicated that the sequence cloned from A. pernyi was the actin A1 promoter . The basic function of the promoter was verified by constructing expression vectors and observing GFP expression. In addition, real-time polymerase chain reaction revealed a strong inhibitory element may exist upstream of the TATA box, which downregulated gene expression. The actin A1 promoter is an ideal candidate for use in A. pernyi transgenic systems.

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/

A novel synonymous mutation causing complete skipping of exon 16 in the SLC26A4 gene in a Korean family with hearing loss

IntroductionMutations in PDS (or SLC26A4) cause both Pendred syndrome (PS) and DFNB4, two autosomal recessive disorders that share hearing loss as a common feature. PS and DFNB4 are genetically homogeneous disorders caused by bi-allelic SLC26A4 mutations. Here, we report a novel synonymous mutation (c.1803G>A, p.Lys601Lys), that caused aberrant splicing in two Korean family members who were clinically considered to have DFNB4, along with congenital hearing loss and dilated vestibular aqueducts (DVA). MethodsAfter extracting DNA from whole blood using standard procedures, the 21 exons and flanking introns of SLC26A4 were amplified with PCR. To evaluate the implication of a novel synonymous mutation (c.1803G>A), we used The Berkeley Drosophila Genome Project (BDGP) (http://www.fruitfly.org/) as a splice site prediction program and performed exon trapping analysis. ResultsIn molecular analysis of the 21 exons of SCL26A4, we detected a known splicing mutation (c.919-2A>G, heterozygote) and a novel variant (c.1803G>A, heterozygote) in the patients (II-1 and II-2). According to in silico analysis, the novel variant (c.1803G>A) affects canonical splice donor nucleotide positioning. To define the transcript level effects of this novel 1803G>A variant, we performed exon trapping and confirmed that exon 16 is completely skipped in this variant type. ConclusionWe report a novel synonymous mutation (c.1803G>A) causing complete exon 16 skipping in the SLC26A4 gene in two Korean family members with hearing loss. This is the first case of a synonymous SNP (c.1803G>A) affecting vestibulocochlear organs through altering splicing accuracy by causing a complete skipping of exon 16. An important issue raised by this study is that synonymous mutations that have been previously ignored in clinical diagnoses must now be considered as potential pathogenic mutations.

Joint stage recognition and anatomical annotation of drosophila gene expression patterns

Motivation: Staining the mRNA of a gene via in situ hybridization (ISH) during the development of a Drosophila melanogaster embryo delivers the detailed spatio-temporal patterns of the gene expression. Many related biological problems such as the detection of co-expressed genes, co-regulated genes and transcription factor binding motifs rely heavily on the analysis of these image patterns. To provide the text-based pattern searching for facilitating related biological studies, the images in the Berkeley Drosophila Genome Project (BDGP) study are annotated with developmental stage term and anatomical ontology terms manually by domain experts. Due to the rapid increase in the number of such images and the inevitable bias annotations by human curators, it is necessary to develop an automatic method to recognize the developmental stage and annotate anatomical terms.Results: In this article, we propose a novel computational model for jointly stage classification and anatomical terms annotation of Drosophila gene expression patterns. We propose a novel Tri-Relational Graph (TG) model that comprises the data graph, anatomical term graph, developmental stage term graph, and connect them by two additional graphs induced from stage or annotation label assignments. Upon the TG model, we introduce a Preferential Random Walk (PRW) method to jointly recognize developmental stage and annotate anatomical terms by utilizing the interrelations between two tasks. The experimental results on two refined BDGP datasets demonstrate that our joint learning method can achieve superior prediction results on both tasks than the state-of-the-art methods.Availability: http://ranger.uta.edu/%7eheng/Drosophila/Contact: heng@uta.edu

Trachealess (Trh) regulates all tracheal genes during Drosophila embryogenesis

The Drosophila trachea is a branched tubular epithelia that transports oxygen and other gases. trachealess ( trh), which encodes a bHLH-PAS transcription factor, is among the first genes to be expressed in the cells that will form the trachea. In the absence of trh, tracheal cells fail to invaginate to form tubes and remain on the embryo surface. Expression of many tracheal-specific genes depends on trh, but all of the known targets have relatively minor phenotypes compared to loss of trh, suggesting that there are additional targets. To identify uncharacterized transcriptional targets of Trh and to further understand the role of Trh in embryonic tracheal formation, we performed an in situ hybridization screen using a library of ~ 100 tracheal-expressed genes identified by the Berkeley Drosophila Genome Project (BDGP). Surprisingly, expression of every tracheal gene we tested was dependent on Trh, suggesting a major role for Trh in activation and maintenance of tracheal gene expression. A re-examination of the interdependence of the known early-expressed transcription factors, including trh, ventral veinless ( vvl) and knirps/ knirps-related ( kni/ knrl), suggests a new model for how gene expression is controlled in the trachea, with trh regulating expression of vvl and kni, but not vice versa. A pilot screen for the targets of Vvl and Kni/Knrl revealed that Vvl and Kni have only minor roles compared to Trh. Finally, genome-wide microarray experiments identified additional Trh targets and revealed that a variety of biological processes are affected by the loss of trh.

RNAi Interference by dsRNA Injection into <em>Drosophila</em> Embryos

Genetic screening is one of the most powerful methods available for gaining insights into complex biological process (1). Over the years many improvements and tools for genetic manipulation have become available in Drosophila (2). Soon after the initial discovery by Frie and Mello (3) that double stranded RNA can be used to knockdown the activity of individual genes in Caenorhabditis elegans, RNA interference (RNAi) was shown to provide a powerful reverse genetic approach to analyze gene functions in Drosophila organ development (4, 5). Many organs, including lung, kidney, liver, and vascular system, are composed of branched tubular networks that transport vital fluids or gases (6, 7). The analysis of Drosophila tracheal formation provides an excellent model system to study the morphogenesis of other tubular organs (8). The Berkeley Drosophila genome project has revealed hundreds of genes that are expressed in the tracheal system. To study the molecular and cellular mechanism of tube formation, the challenge is to understand the roles of these genes in tracheal development. Here, we described a detailed method of dsRNA injection into Drosophila embryo to knockdown individual gene expression. We successfully knocked down endogenous dysfusion(dys) gene expression by dsRNA injection. Dys is a bHLH-PAS protein expressed in tracheal fusion cells, and it is required for tracheal branch fusion (9, 10). dys-RNAi completely eliminated dys expression and resulted in tracheal fusion defect. This relatively simple method provides a tool to identify genes required for tissue and organ development in Drosophila.

RNAi Interference by dsRNA Injection into <em>Drosophila</em> Embryos

Genetic screening is one of the most powerful methods available for gaining insights into complex biological process 1. Over the years many improvements and tools for genetic manipulation have become available in Drosophila 2. Soon after the initial discovery by Frie and Mello 3 that double stranded RNA can be used to knockdown the activity of individual genes in Caenorhabditis elegans, RNA interference (RNAi) was shown to provide a powerful reverse genetic approach to analyze gene functions in Drosophila organ development 4, 5. Many organs, including lung, kidney, liver, and vascular system, are composed of branched tubular networks that transport vital fluids or gases 6, 7. The analysis of Drosophila tracheal formation provides an excellent model system to study the morphogenesis of other tubular organs 8. The Berkeley Drosophila genome project has revealed hundreds of genes that are expressed in the tracheal system. To study the molecular and cellular mechanism of tube formation, the challenge is to understand the roles of these genes in tracheal development. Here, we described a detailed method of dsRNA injection into Drosophila embryo to knockdown individual gene expression. We successfully knocked down endogenous dysfusion(dys) gene expression by dsRNA injection. Dys is a bHLH-PAS protein expressed in tracheal fusion cells, and it is required for tracheal branch fusion 9, 10. dys-RNAi completely eliminated dys expression and resulted in tracheal fusion defect. This relatively simple method provides a tool to identify genes requried for tissure and organ development in Drosophila.

Mutation screening of RAD51C in male breast cancer patients

We read with great interest the paper by Akbari and colleagues [1] in a recent issue of Breast Cancer Research. The authors reported on the absence of RAD51C mutations in 454 patients with BRCA1/2-negative familial breast cancer/ovarian cancer (BC/OC). In the initial report by Meindl and colleagues [2], RAD51C mutations were identified in 6 out of 480 patients with BRCA1/2-negative familial BC/OC. Interestingly, on the basis of histopathologic features, including intermediate grade (G2), estrogen receptor-positive (ER+), progesterone receptor-positive (PR+), and HER2-negative (HER2-) expression, RAD51C-associated BCs were found to be similar to BRCA2-associated BCs [2]. BRCA2 is known to play a significant role in male BC (MBC); however, no occurrence of MBC was observed in the six RAD51C families described [2]. To investigate the role of RAD51C in MBC, we screened for RAD51C mutations in 97 MBC patients selected from our population-based series of 126 cases because they were previously found negative for BRCA1/2, CHEK2, PALB2, and BRIP1 mutations [3,4]. Notably, 25.8% of cases showed a positive first-degree family history of BC or OC or both. The majority of MBCs were invasive ductal carcinomas (74.5%), G2 (53.5%), ER+ (90.7%), PR+ (82.6%), or HER2- (85.4%). Overall, 66% of the MBCs showed an ER+/PR+/HER2- phenotype. All patients provided informed consent to the study. We carried out mutation screening of the nine exons and intron/exon boundaries of RAD51C by high resolution melting (HRM) analysis, a rapid closed-tube mutation scanning method with high sensitivity and specificity. Cases displaying abnormal profiles were evaluated by direct sequencing. Primers are available upon request. We found no truncating RAD51C mutations. We identified a novel intronic variant, IVS3 c.738-16G>T, in 1 out of 97 MBCs (1%). By in silico analysis, performed with Splice Site Prediction (BDGP, Berkeley Drosophila Genome Project, Lawrence Berkeley National Laboratory, Berkeley, CA, USA) and NetGene 2 Server software (CBSA, Center for Biological Sequence Analysis, Technical University of Denmark, Lyngby, Denmark), the IVS3 c.738-16G>T variant is predicted not to affect splicing. This variant was also identified in 2 out of 173 (1.2%) population controls examined. We also found a neutral polymorphic intronic variant, IVS6 c.904+34T>C (rs28363318), in 16 out of 97 (16.5%) MBCs. Overall, our results, which are based on a relatively large MBC series, are consistent with the findings by Akbari and colleagues [1] and with data on 92 patients with hereditary gynecological cancer in which no deleterious RAD51C mutations were identified [5] and would suggest that the impact of RAD51C mutations on BC predisposition might be more limited than initially reported. In conclusion, we found no evidence that RAD51C mutations may contribute to MBC susceptibility. Further studies on larger MBC series are needed to confirm our findings.

CHN1 mutations are not a common cause of sporadic Duane's retraction syndrome

<i>CHN1</i> mutations are not a common cause of sporadic Duane's retraction syndrome

A bag-of-words approach for Drosophila gene expression pattern annotation

BackgroundDrosophila gene expression pattern images document the spatiotemporal dynamics of gene expression during embryogenesis. A comparative analysis of these images could provide a fundamentally important way for studying the regulatory networks governing development. To facilitate pattern comparison and searching, groups of images in the Berkeley Drosophila Genome Project (BDGP) high-throughput study were annotated with a variable number of anatomical terms manually using a controlled vocabulary. Considering that the number of available images is rapidly increasing, it is imperative to design computational methods to automate this task.ResultsWe present a computational method to annotate gene expression pattern images automatically. The proposed method uses the bag-of-words scheme to utilize the existing information on pattern annotation and annotates images using a model that exploits correlations among terms. The proposed method can annotate images individually or in groups (e.g., according to the developmental stage). In addition, the proposed method can integrate information from different two-dimensional views of embryos. Results on embryonic patterns from BDGP data demonstrate that our method significantly outperforms other methods.ConclusionThe proposed bag-of-words scheme is effective in representing a set of annotations assigned to a group of images, and the model employed to annotate images successfully captures the correlations among different controlled vocabulary terms. The integration of existing annotation information from multiple embryonic views improves annotation performance.

Recovery of DNA Sequences Flanking P-Element Insertions in Drosophila: Inverse PCR and Plasmid Rescue

The Drosophila melanogaster P-transposable element is a powerful and widely used research tool. Sequences flanking the P-element can be recovered and the site of insertion can be mapped to the nucleotide, to connect the genetic and physical maps and facilitate molecular analysis of the gene of interest. The Berkeley Drosophila Genome Project (BDGP) has assembled a well-characterized collection of lethal mutations induced by single P-element insertions generated by a number of laboratories. The genomic DNA sequences adjacent to these insertions have been recovered by either plasmid rescue or inverse polymerase chain reaction (PCR). The combination of a complete genomic DNA sequence and relatively fast and easy molecular methods for mapping P-element insertion sites to the nucleotide enhances the use of P-elements as tools in Drosophila research. This protocol provides detailed procedures for isolating DNA flanking P-element insertions.

The DHHC Palmitoyltransferase Approximated Regulates Fat Signaling and Dachs Localization and Activity

Signaling via the large protocadherin Fat (Ft), regulated in part by its binding partner Dachsous (Ds) and the Golgi-resident kinase Four-jointed (Fj), is required for a variety of developmental functions in Drosophila. Ft and, to a lesser extent, Ds suppress overgrowth of the imaginal discs from which appendages develop and regulate the Hippo pathway [1-5] (reviewed in [6]). Ft, Ds, and Fj are also required for normal planar cell polarity (PCP) in the wing, abdomen, and eye and for the normal patterning of appendages, including the spacing of crossveins in the wing and the segmentation of the leg tarsus (reviewed in [7-9]). Ft signaling was recently shown to be negatively regulated by the atypical myosin Dachs [10, 11]. We identify here an additional negative regulator of Ft signaling in growth control, PCP, and appendage patterning, the Approximated (App) protein. We show that App encodes a member of the DHHC family, responsible for the palmitoylation of selected cytoplasmic proteins, and provide evidence that App acts by controlling the normal subcellular localization and activity of Dachs.

Gene structure of Drosophila diaphorase-1: diversity of transcripts in adult males and females, in different organs and at different stages of development

The gene EG:22E5.5 or CG4199 (accession number O77266, Q9W529) from Berkeley Drosophila Genome Project (BDGP) was found using the partial amino acid sequences of three tryptic peptides obtained from purified Drosophila virilis diaphorase-1. This gene is located on the X chromosome at position 2C9-2C10. The structure of the gene reveals three exons and two long introns. Using BDGP, we found six transcripts in this gene. The difference between these transcripts is in their 5' ends; the 3' ends of the six transcripts are identical. Thirty-four ESTs from different cDNA libraries were found, most of them from Schneider L2 cell culture (SH) cDNA library. The transcripts are represented at very low level in the cells of different organs and at different stages of Drosophila development. Using RT-PCR, we obtained five of these transcripts in cDNA samples from female adult flies. However, we could not find any of them in cDNA samples from male adult flies. Moreover, we obtained only the third transcript (CG4199-RC) in the sample of testis from adult flies and the fourth transcript (CG4199-RD) in an embryo sample. None of the other five transcripts were found in the samples of different organs and in the samples obtained at different stages of Drosophila development.

Requirement for a Drosophila E3-Ubiquitin Ligase in Phagocytosis of Apoptotic Cells

Many cells die by apoptosis during animal development. Apoptotic cells are rapidly removed through phagocytosis by their neighbors or by macrophages. To genetically dissect this process, we performed an in vivo screen for genes required for efficient phagocytosis of apoptotic cells by Drosophila macrophages and identified pallbearer (pall), which encodes an F box protein. F box proteins generally provide substrate specificity to Skp Cullin F box (SCF) complexes, acting as E3 ligases that target phosphorylated proteins to ubiquitylation and degradation via the 26S proteasome. We showed that Pallbearer functions in an SCF-dependent manner and provided direct evidence of a role for ubiquitylation and proteasomal degradation in phagocytosis of apoptotic corpses in vivo. This work might further our understanding of the regulation of apoptotic cell engulfment and thus our understanding of innate immunity as a whole.

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.

Molecular Characterization of teflon, a Gene Required for Meiotic Autosome Segregation in Male Drosophila melanogaster

Drosophila melanogaster males lack recombination and have evolved a mechanism of meiotic chromosome segregation that is independent of both the chiasmatic and achiasmatic segregation systems of females. The teflon (tef) gene is specifically required in males for proper segregation of autosomes and provides a genetic tool for understanding recombination-independent mechanisms of pairing and segregation as well as differences in sex chromosome vs. autosome segregation. Here we report on the cloning of the tef gene and the molecular characterization of tef mutations. Rescue experiments using a GAL4-driven pUAS transgene demonstrate that tef corresponds to predicted Berkeley Drosophila Genome Project (BDGP) gene CG8961 and that tef expression is required in the male germ line prior to spermatocyte stage S4. Consistent with this early prophase requirement, expression of tef was found to be independent of regulators of meiotic M phase initiation or progression. The predicted Tef protein contains three C2H2 zinc-finger motifs, one at the amino terminus and two in tandem at the carboxyl terminus. In addition to the zinc-finger motifs, a 44- to 45-bp repeat is conserved in three related Drosophila species. On the basis of these findings, we propose a role for Tef as a bridging molecule that holds autosome bivalents together via heterochromatic connections.