Identifying small RNAs derived from maternal- and somatic-type rRNAs in zebrafish development.

rRNAs are non-coding RNAs present in all prokaryotes and eukaryotes. In eukaryotes there are four rRNAs: 18S, 5.8S, 28S, originating from a common precursor (45S), and 5S. We have recently discovered the existence of two distinct developmental types of rRNA: a maternal-type, present in eggs and a somatic-type, expressed in adult tissues. Lately, next-generation sequencing has allowed the discovery of new small-RNAs deriving from longer non-coding RNAs, including small-RNAs from rRNAs (srRNAs). Here, we systemically investigated srRNAs of maternal- or somatic-type 18S, 5.8S, 28S, with small-RNAseq from many zebrafish developmental stages. We identified new srRNAs for each rRNA. For 5.8S, we found srRNA consisting of the 5' or 3' halves, with only the latter having different sequence for the maternal- and somatic-types. For 18S, we discovered 21 nt srRNA from the 5' end of the 18S rRNA with a striking resemblance to microRNAs; as it is likely processed from a stem-loop precursor and present in human and mouse Argonaute-complexed small-RNA. For 28S, an abundant 80 nt srRNA from the 3' end of the 28S rRNA was found. The expression levels during embryogenesis of these srRNA indicate they are not generated from rRNA degradation and might have a role in the zebrafish development.

Expression of distinct maternal and somatic 5.8S, 18S, and 28S rRNA types during zebrafish development.

There is mounting evidence that the ribosome is not a static translation machinery, but a cell-specific, adaptive system. Ribosomal variations have mostly been studied at the protein level, even though the essential transcriptional functions are primarily performed by rRNAs. At the RNA level, oocyte-specific 5S rRNAs are long known for Xenopus. Recently, we described for zebrafish a similar system in which the sole maternal-type 5S rRNA present in eggs is replaced completely during embryonic development by a somatic-type. Here, we report the discovery of an analogous system for the 45S rDNA elements: 5.8S, 18S, and 28S. The maternal-type 5.8S, 18S, and 28S rRNA sequences differ substantially from those of the somatic-type, plus the maternal-type rRNAs are also replaced by the somatic-type rRNAs during embryogenesis. We discuss the structural and functional implications of the observed sequence differences with respect to the translational functions of the 5.8S, 18S, and 28S rRNA elements. Finally, in silico evidence suggests that expansion segments (ES) in 18S rRNA, previously implicated in ribosome-mRNA interaction, may have a preference for interacting with specific mRNA genes. Taken together, our findings indicate that two distinct types of ribosomes exist in zebrafish during development, each likely conducting the translation machinery in a unique way.

Transcriptome dynamics in early zebrafish embryogenesis determined by high-resolution time course analysis of 180 successive, individual zebrafish embryos.

BACKGROUND: Recently, much progress has been made in the field of gene-expression in early embryogenesis. However, the dynamic behaviour of transcriptomes in individual embryos has hardly been studied yet and the time points at which pools of embryos are collected are usually still quite far apart. Here, we present a high-resolution gene-expression time series with 180 individual zebrafish embryos, obtained from nine different spawns, developmentally ordered and profiled from late blastula to mid-gastrula stage. On average one embryo per minute was analysed. The focus was on identification and description of the transcriptome dynamics of the expressed genes in this embryonic stage, rather than to biologically interpret profiles in cellular processes and pathways. RESULTS: In the late blastula to mid-gastrula stage, we found 6,734 genes being expressed with low variability and rather gradual changes. Ten types of dynamic behaviour were defined, such as genes with continuously increasing or decreasing expression, and all expressed genes were grouped into these types. Also, the exact expression starting and stopping points of several hundred genes during this developmental period could be pinpointed. Although the resolution of the experiment was so high, that we were able to clearly identify four known oscillating genes, no genes were observed with a peaking expression. Additionally, several genes showed expression at two or three distinct levels that strongly related to the spawn an embryo originated from. CONCLUSION: Our unique experimental set-up of whole-transcriptome analysis of 180 individual embryos, provided an unparalleled in-depth insight into the dynamics of early zebrafish embryogenesis. The existence of a tightly regulated embryonic transcriptome program, even between individuals from different spawns is shown. We have made the expression profile of all genes available for domain experts. The fact that we were able to separate the different spawns by their gene-expression variance over all expressed genes, underlines the importance of spawn specificity, as well as the unexpectedly tight gene-expression regulation in early zebrafish embryogenesis.

Linking maternal and somatic 5S rRNA types with different sequence-specific non-LTR retrotransposons.

5S rRNA is a ribosomal core component, transcribed from many gene copies organized in genomic repeats. Some eukaryotic species have two 5S rRNA types defined by their predominant expression in oogenesis or adult tissue. Our next-generation sequencing study on zebrafish egg, embryo, and adult tissue identified maternal-type 5S rRNA that is exclusively accumulated during oogenesis, replaced throughout the embryogenesis by a somatic-type, and thus virtually absent in adult somatic tissue. The maternal-type 5S rDNA contains several thousands of gene copies on chromosome 4 in tandem repeats with small intergenic regions, whereas the somatic-type is present in only 12 gene copies on chromosome 18 with large intergenic regions. The nine-nucleotide variation between the two 5S rRNA types likely affects TFIII binding and riboprotein L5 binding, probably leading to storage of maternal-type rRNA. Remarkably, these sequence differences are located exactly at the sequence-specific target site for genome integration by the 5S rRNA-specific Mutsu retrotransposon family. Thus, we could define maternal- and somatic-type MutsuDr subfamilies. Furthermore, we identified four additional maternal-type and two new somatic-type MutsuDr subfamilies, each with their own target sequence. This target-site specificity, frequently intact maternal-type retrotransposon elements, plus specific presence of Mutsu retrotransposon RNA and piRNA in egg and adult tissue, suggest an involvement of retrotransposons in achieving the differential copy number of the two types of 5S rDNA loci.

Transcriptome data on maternal RNA of 24 individual zebrafish eggs from five sibling mothers.

Maternal mRNA that is present in the mature oocyte plays an important role in the proper development of the early embryo. To elucidate the role of the maternal transcriptome we recently reported a microarray study on individual zebrafish eggs from five different clutches from sibling mothers and showed differences in maternal RNA abundance between and within clutches, "Mother-specific signature in the maternal transcriptome composition of mature, unfertilized Eggs" [1]. Here we provide in detail the applied preprocessing method as well as the R-code to identify expressed and non-expressed genes in the associated transcriptome dataset. Additionally, we provide a website that allows a researcher to search for the expression of their gene of interest in this experiment.

Confounding Factors in the Transcriptome Analysis of an In-Vivo Exposure Experiment.

CONFOUNDING FACTORS: In transcriptomics experimentation, confounding factors frequently exist alongside the intended experimental factors and can severely influence the outcome of a transcriptome analysis. Confounding factors are regularly discussed in methodological literature, but their actual, practical impact on the outcome and interpretation of transcriptomics experiments is, to our knowledge, not documented. For instance, in-vivo experimental factors; like Individual, Sample-Composition and Time-of-Day are potentially formidable confounding factors. To study these confounding factors, we designed an extensive in-vivo transcriptome experiment (n = 264) with UVR exposure of murine skin containing six consecutive samples from each individual mouse (n = 64). ANALYSIS APPROACH: Evaluation of the confounding factors: Sample-Composition, Time-of-Day, Handling-Stress, and Individual-Mouse resulted in the identification of many genes that were affected by them. These genes sometimes showed over 30-fold expression differences. The most prominent confounding factor was Sample-Composition caused by mouse-dependent skin composition differences, sampling variation and/or influx/efflux of mobile cells. Although we can only evaluate these effects for known cell type specifically expressed genes in our complex heterogeneous samples, it is clear that the observed variations also affect the cumulative expression levels of many other non-cell-type-specific genes. ANOVA: ANOVA analysis can only attempt to neutralize the effects of the well-defined confounding factors, such as Individual-Mouse, on the experimental factors UV-Dose and Recovery-Time. Also, by definition, ANOVA only yields reproducible gene-expression differences, but we found that these differences were very small compared to the fold changes induced by the confounding factors, questioning the biological relevance of these ANOVA-detected differences. Furthermore, it turned out that many of the differentially expressed genes found by ANOVA were also present in the gene clusters associated with the confounding factors. CONCLUSION: Hence our overall conclusion is that confounding factors have a major impact on the outcome of in-vivo transcriptomics experiments. Thus the set-up, analysis, and interpretation of such experiments should be approached with the utmost prudence.

Mother-Specific Signature in the Maternal Transcriptome Composition of Mature, Unfertilized Zebrafish Eggs.

Maternal mRNA present in mature oocytes plays an important role in the proper development of the early embryo. As the composition of the maternal transcriptome in general has been studied with pooled mature eggs, potential differences between individual eggs are unknown. Here we present a transcriptome study on individual zebrafish eggs from clutches of five mothers in which we focus on the differences in maternal mRNA abundance per gene between and within clutches. To minimize technical interference, we used mature, unfertilized eggs from siblings. About half of the number of analyzed genes was found to be expressed as maternal RNA. The expressed and non-expressed genes showed that maternal mRNA accumulation is a non-random process, as it is related to specific biological pathways and processes relevant in early embryogenesis. Moreover, it turned out that overall the composition of the maternal transcriptome is tightly regulated as about half of the expressed genes display a less than twofold expression range between the observed minimum and maximum expression values of a gene in the experiment. Even more, the maximum gene-expression difference within clutches is for 88% of the expressed genes lower than twofold. This means that expression differences observed in maternally expressed genes are primarily caused by differences between mothers, with only limited variability between eggs from the same mother. This was underlined by the fact that 99% of the expressed genes were found to be differentially expressed between any of the mothers in an ANOVA test. Furthermore, linking chromosome location, transcription factor binding sites, and miRNA target sites of the genes in clusters of distinct and unique mother-specific gene-expression, suggest biological relevance of the mother-specific signatures in the maternal transcriptome composition. Altogether, the maternal transcriptome composition of mature zebrafish oocytes seems to be tightly regulated with a distinct mother-specific signature.

WRKY40 and WRKY6 act downstream of the green leaf volatile E-2-hexenal in Arabidopsis.

Plants are known to be responsive to volatiles, but knowledge about the molecular players involved in transducing their perception remains scarce. We study the response of Arabidopsis thaliana to E-2-hexenal, one of the green leaf volatiles (GLV) that is produced upon wounding, herbivory or infection with pathogens. We have taken a transcriptomics approach to identify genes that are induced by E-2-hexenal, but not by defence hormones or other GLVs. Furthermore, by studying the promoters of early E-2-hexenal-induced genes we determined that the only statistically enriched cis-element was the W-box motif. Since members of the plant-specific family of WRKY transcription factors act in trans on this cis-element, we focused on WRKY6, 40 and 53 that were most strongly induced by E-2-hexenal. Root elongation of Arabidopsis seedlings of the wrky40 wrky6 double mutant was much less inhibited than in wt plants, similar to the E-2-hexenal-responsive mutant her1, which is perturbed in γ-amino butyric acid (GABA) metabolism. The induction of several of the E-2-hexenal-specific genes was much higher in the wrky40, wrky6 or wrky40 wrky6 mutants, including GAD4, a glutamate decarboxylase that catalyzes the formation of GABA from glutamate. In conclusion, WRKY6 and 40 seem to act as important players transducing E-2-hexenal perception.

Valuable lessons-learned in transcriptomics experimentation.

We have collected several valuable lessons that will help improve transcriptomics experimentation. These lessons relate to experiment design, execution, and analysis. The cautions, but also the pointers, may help biologists avoid common pitfalls in transcriptomics experimentation and achieve better results with their transcriptome studies.

Improving small RNA-seq by using a synthetic spike-in set for size-range quality control together with a set for data normalization.

There is an increasing interest in complementing RNA-seq experiments with small-RNA (sRNA) expression data to obtain a comprehensive view of a transcriptome. Currently, two main experimental challenges concerning sRNA-seq exist: how to check the size distribution of isolated sRNAs, given the sensitive size-selection steps in the protocol; and how to normalize data between samples, given the low complexity of sRNA types. We here present two separate sets of synthetic RNA spike-ins for monitoring size-selection and for performing data normalization in sRNA-seq. The size-range quality control (SRQC) spike-in set, consisting of 11 oligoribonucleotides (10-70 nucleotides), was tested by intentionally altering the size-selection protocol and verified via several comparative experiments. We demonstrate that the SRQC set is useful to reproducibly track down biases in the size-selection in sRNA-seq. The external reference for data-normalization (ERDN) spike-in set, consisting of 19 oligoribonucleotides, was developed for sample-to-sample normalization in differential-expression analysis of sRNA-seq data. Testing and applying the ERDN set showed that it can reproducibly detect differential expression over a dynamic range of 2(18). Hence, biological variation in sRNA composition and content between samples is preserved while technical variation is effectively minimized. Together, both spike-in sets can significantly improve the technical reproducibility of sRNA-seq.

A range finding protocol to support design for transcriptomics experimentation: examples of in-vitro and in-vivo murine UV exposure.

In transcriptomics research, design for experimentation by carefully considering biological, technological, practical and statistical aspects is very important, because the experimental design space is essentially limitless. Usually, the ranges of variable biological parameters of the design space are based on common practices and in turn on phenotypic endpoints. However, specific sub-cellular processes might only be partially reflected by phenotypic endpoints or outside the associated parameter range. Here, we provide a generic protocol for range finding in design for transcriptomics experimentation based on small-scale gene-expression experiments to help in the search for the right location in the design space by analyzing the activity of already known genes of relevant molecular mechanisms. Two examples illustrate the applicability: in-vitro UV-C exposure of mouse embryonic fibroblasts and in-vivo UV-B exposure of mouse skin. Our pragmatic approach is based on: framing a specific biological question and associated gene-set, performing a wide-ranged experiment without replication, eliminating potentially non-relevant genes, and determining the experimental 'sweet spot' by gene-set enrichment plus dose-response correlation analysis. Examination of many cellular processes that are related to UV response, such as DNA repair and cell-cycle arrest, revealed that basically each cellular (sub-) process is active at its own specific spot(s) in the experimental design space. Hence, the use of range finding, based on an affordable protocol like this, enables researchers to conveniently identify the 'sweet spot' for their cellular process of interest in an experimental design space and might have far-reaching implications for experimental standardization.

Absence/presence calling in microarray-based CGH experiments with non-model organisms.

Structural variations in genomes are commonly studied by (micro)array-based comparative genomic hybridization. The data analysis methods to infer copy number variation in model organisms (human, mouse) are established. In principle, the procedures are based on signal ratios between test and reference samples and the order of the probe targets in the genome. These procedures are less applicable to experiments with non-model organisms, which frequently comprise non-sequenced genomes with an unknown order of probe targets. We therefore present an additional analysis approach, which does not depend on the structural information of a reference genome, and quantifies the presence or absence of a probe target in an unknown genome. The principle is that intensity values of target probes are compared with the intensities of negative-control probes and positive-control probes from a control hybridization, to determine if a probe target is absent or present. In a test, analyzing the genome content of a known bacterial strain: Staphylococcus aureus MRSA252, this approach proved to be successful, demonstrated by receiver operating characteristic area under the curve values larger than 0.9995. We show its usability in various applications, such as comparing genome content and validating next-generation sequencing reads from eukaryotic non-model organisms.

Transcriptome analysis of Traf6 function in the innate immune response of zebrafish embryos.

TRAF6 is a key player at the cross-roads of development and immunity. The analysis of its in vivo molecular function is a great challenge since severe developmental defects and early lethality caused by Traf6 deficiency in knock-out mice interfere with analyses of the immune response. In this study we have used a new strategy to analyze the function of Traf6 in a zebrafish-Salmonella infectious disease model. In our approach the effect of a Traf6 translation-blocking morpholino was titrated down to avoid developmental defects and the response to infection under these conditions was studied using the combination of microarray analysis and whole transcriptome deep sequencing. Transcriptome profiling of the traf6 knock-down allowed the identification of a gene set whose responsiveness during infection is highly dependent on Traf6. Expression trend analysis based on the resulting datasets identified nine clusters of genes with characteristic transcription response profiles, demonstrating Traf6 has a dynamic role as a positive and negative regulator. Among the Traf6-dependent genes was a large set of well known anti-microbial and inflammatory genes. Additionally, we identified several genes which were not previously linked to a response to microbial infection, such as the fertility hormone gene gnrh2 and the DNA-damage regulated autophagy modulator 1 gene dram1. With the use of the zebrafish embryo model we have now analyzed the in vivo function of Traf6 in the innate immune response without interference of adaptive immunity.

Integrating heterogeneous sequence information for transcriptome-wide microarray design; a Zebrafish example.

BACKGROUND: A complete gene-expression microarray should preferably detect all genomic sequences that can be expressed as RNA in an organism, i.e. the transcriptome. However, our knowledge of a transcriptome of any organism still is incomplete and transcriptome information is continuously being updated. Here, we present a strategy to integrate heterogeneous sequence information that can be used as input for an up-to-date microarray design. FINDINGS: Our algorithm consists of four steps. In the first step transcripts from different resources are grouped into Transcription Clusters (TCs) by looking at the similarity of all transcripts. TCs are groups of transcripts with a similar length. If a transcript is much smaller than a TC to which it is highly similar, it will be annotated as a subsequence of that TC and is used for probe design only if the probe designed for the TC does not query the subsequence. Secondly, all TCs are mapped to a genome assembly and gene information is added to the design. Thirdly TC members are ranked according to their trustworthiness and the most reliable sequence is used for the probe design. The last step is the actual array design. We have used this strategy to build an up-to-date zebrafish microarray. CONCLUSIONS: With our strategy and the software developed, it is possible to use a set of heterogeneous transcript resources for microarray design, reduce the number of candidate target sequences on which the design is based and reduce redundancy. By changing the parameters in the procedure it is possible to control the similarity within the TCs and thus the amount of candidate sequences for the design. The annotation of the microarray is carried out simultaneously with the design.

RNA isolation method for single embryo transcriptome analysis in zebrafish.

BACKGROUND: Transcriptome analysis during embryogenesis usually requires pooling of embryos to obtain sufficient RNA. Hence, the measured levels of gene-expression represent the average mRNA levels of pooled samples and the biological variation among individuals is confounded. This can irreversibly reduce the robustness, resolution, or expressiveness of the experiment. Therefore, we developed a robust method to isolate abundant high-quality RNA from individual embryos to perform single embryo transcriptome analyses using zebrafish as a model organism. Available methods for embryonic zebrafish RNA isolation minimally utilize ten embryos. Further downscaling of these methods to one embryo is practically not feasible. FINDINGS: We developed a single embryo RNA extraction method based on sample homogenization in liquid nitrogen, RNA extraction with phenol and column purification. Evaluation of this method showed that: the quality of the RNA was very good with an average RIN value of 8.3-8.9; the yield was always >/= 200 ng RNA per embryo; the method was applicable to all stages of zebrafish embryogenesis; the success rate was almost 100%; and the extracted RNA performed excellent in microarray experiments in that the technical variation was much lower than the biological variation. CONCLUSIONS: Presented is a high-quality, robust RNA isolation method. Obtaining sufficient RNA from single embryos eliminates the necessity of sample pooling and its associated drawbacks. Although our RNA isolation method has been setup for transcriptome analysis in zebrafish, it can also be used for other model systems and other applications like (q)PCR and transcriptome sequencing.

Serious complications in gene-expression studies with stress perturbation: An example of UV-exposed p53-mutant mouse embryonic fibroblasts.

Reanalysis of our UV study of p53-mutant mouse embryonic fibroblasts revealed an intriguing orchestration of massive transcriptome responses. However, close scrutiny of the data uncovered an affected mRNA/rRNA ratio, effectively inhibiting valid data analysis. UV-dose range-finding showed low-dose UV specific- and high-dose stress-related responses, which represent a plea for UV dose range-finding in experimental design.

SigWinR; the SigWin-detector updated and ported to R.

BACKGROUND: Our SigWin-detector discovers significantly enriched windows of (genomic) elements in any sequence of values (genes or other genomic elements in a DNA sequence) in a fast and reproducible way. However, since it is grid based, only (life) scientists with access to the grid can use this tool. Therefore and on request, we have developed the SigWinR package which makes the SigWin-detector available to a much wider audience. At the same time, we have introduced several improvements to its algorithm as well as its functionality, based on the feedback of SigWin-detector end users. FINDINGS: To allow usage of the SigWin-detector on a desktop computer, we have rewritten it as a package for R: SigWinR. R is a free and widely used multi platform software environment for statistical computing and graphics. The package can be installed and used on all platforms for which R is available. The improvements involve: a visualization of the input-sequence values supporting the interpretation of Ridgeograms; a visualization allowing for an easy interpretation of enriched or depleted regions in the sequence using windows of pre-defined size; an option that allows the analysis of circular sequences, which results in rectangular Ridgeograms; an application to identify regions of co-altered gene expression (ROCAGEs) with a real-life biological use-case; adaptation of the algorithm to allow analysis of non-regularly sampled data using a constant window size in physical space without resampling the data. To achieve this, support for analysis of windows with an even number of elements was added. CONCLUSION: By porting the SigWin-detector as an R package, SigWinR, improving its algorithm and functionality combined with adequate performance, we have made SigWin-detector more useful as well as more easily accessible to scientists without a grid infrastructure.

Using R in Taverna: RShell v1.2.

BACKGROUND: R is the statistical language commonly used by many life scientists in (omics) data analysis. At the same time, these complex analyses benefit from a workflow approach, such as used by the open source workflow management system Taverna. However, Taverna had limited support for R, because it supported just a few data types and only a single output. Also, there was no support for graphical output and persistent sessions. Altogether this made using R in Taverna impractical. FINDINGS: We have developed an R plugin for Taverna: RShell, which provides R functionality within workflows designed in Taverna. In order to fully support the R language, our RShell plugin directly uses the R interpreter. The RShell plugin consists of a Taverna processor for R scripts and an RShell Session Manager that communicates with the R server. We made the RShell processor highly configurable allowing the user to define multiple inputs and outputs. Also, various data types are supported, such as strings, numeric data and images. To limit data transport between multiple RShell processors, the RShell plugin also supports persistent sessions. Here, we will describe the architecture of RShell and the new features that are introduced in version 1.2, i.e.: i) Support for R up to and including R version 2.9; ii) Support for persistent sessions to limit data transfer; iii) Support for vector graphics output through PDF; iv)Syntax highlighting of the R code; v) Improved usability through fewer port types.Our new RShell processor is backwards compatible with workflows that use older versions of the RShell processor. We demonstrate the value of the RShell processor by a use-case workflow that maps oligonucleotide probes designed with DNA sequence information from Vega onto the Ensembl genome assembly. CONCLUSION: Our RShell plugin enables Taverna users to employ R scripts within their workflows in a highly configurable way.

OligoRAP - an Oligo Re-Annotation Pipeline to improve annotation and estimate target specificity.

BACKGROUND: High throughput gene expression studies using oligonucleotide microarrays depend on the specificity of each oligonucleotide (oligo or probe) for its target gene. However, target specific probes can only be designed when a reference genome of the species at hand were completely sequenced, when this genome were completely annotated and when the genetic variation of the sampled individuals were completely known. Unfortunately there is not a single species for which such a complete data set is available. Therefore, it is important that probe annotation can be updated frequently for optimal interpretation of microarray experiments. RESULTS: In this paper we present OligoRAP, a pipeline to automatically update the annotation of oligo libraries and estimate oligo target specificity. OligoRAP uses a reference genome assembly with Ensembl and Entrez Gene annotation supplemented with a set of unmapped transcripts derived from RefSeq and UniGene to handle assembly gaps. OligoRAP produces alignments of each oligo with the reference assembly as well as with unmapped transcripts. These alignments are re-mapped to the annotation sources, which results in a concise, as complete as possible and up-to-date annotation of the oligo library. The building blocks of this pipeline are BioMoby web services creating a highly modular and distributed system with a robust, remote programmatic interface.OligoRAP was used to update the annotation for a subset of 791 oligos from the ARK-Genomics 20 K chicken array, which were selected as starting material for the oligo annotation session of the EADGENE/SABRE Post-analysis workshop. Based on the updated annotation about one third of these oligos is problematic with regard to target specificity. In addition, the accession numbers or ids the oligos were originally designed for no longer exist in the updated annotation for almost half of the oligos. CONCLUSION: As microarrays are designed on incomplete data, it is important to update probe annotation and check target specificity regularly. OligoRAP provides both and due to its design based on BioMoby web services it can easily be embedded as an oligo annotation engine in customised applications for microarray data analysis. The dramatic difference in updated annotation and target specificity for the ARK-Genomics 20 K chicken array as compared to the original data emphasises the need for regular updates.

Salvaging Affymetrix probes after probe-level re-annotation.

BACKGROUND: Affymetrix GeneChips can be re-annotated at the probe-level by breaking up the original probe-sets and recomposing new probe-sets based on up-to-date genomic knowledge, such as available in Entrez Gene. This results in custom Chip Description Files (CDF). Using these custom CDFs improves the quality of the data and thus the results of related gene expression studies. However, 44-71% of the probes on a GeneChip are lost in this re-annotation process. Although generally aimed at less known genes, losing these probes obviously means a substantial loss of expensive experiment data. Biologists are therefore very reluctant to adopt this approach. FINDINGS: We aimed to re-introduce the non-affected Affymetrix probe-sets after these re-annotation procedures. For this, we developed an algorithm (CDF-Merger) and applied it to standard Affymetrix CDFs and custom Brainarray CDFs to obtain Hybrid CDFs. Thus, salvaging lost Affymetrix probes with our CDF-Merger restored probe content up to 94%. Because the salvaged probes (up to 54% of the probe content on the arrays) represent less-reliable probe-sets, we made the origin of all probe-set definitions traceable, so biologists can choose at any time in their analyses, which subset of probe-sets they want to use. CONCLUSION: The availability of up-to-date Hybrid CDFs plus R environment allows for easy implementation of our approach.