PiGx: reproducible genomics analysis pipelines with GNU Guix

In bioinformatics, as well as other computationally intensive research fields, there is a need for workflows that can reliably produce consistent output, from known sources, independent of the software environment or configuration settings of the machine on which they are executed. Indeed, this is essential for controlled comparison between different observations and for the wider dissemination of workflows. However, providing this type of reproducibility and traceability is often complicated by the need to accommodate the myriad dependencies included in a larger body of software, each of which generally comes in various versions. Moreover, in many fields (bioinformatics being a prime example), these versions are subject to continual change due to rapidly evolving technologies, further complicating problems related to reproducibility. Here, we propose a principled approach for building analysis pipelines and managing their dependencies with GNU Guix. As a case study to demonstrate the utility of our approach, we present a set of highly reproducible pipelines called PiGx for the analysis of RNA sequencing, chromatin immunoprecipitation sequencing, bisulfite-treated DNA sequencing, and single-cell resolution RNA sequencing. All pipelines process raw experimental data and generate reports containing publication-ready plots and figures, with interactive report elements and standard observables. Users may install these highly reproducible packages and apply them to their own datasets without any special computational expertise beyond the use of the command line. We hope such a toolkit will provide immediate benefit to laboratory workers wishing to process their own datasets or bioinformaticians seeking to automate all, or parts of, their analyses. In the long term, we hope our approach to reproducibility will serve as a blueprint for reproducible workflows in other areas. Our pipelines, along with their corresponding documentation and sample reports, are available at http://bioinformatics.mdc-berlin.de/pigx.

Development and validation of a nomogram to predict recurrence and melanoma-specific mortality in patients with negative sentinel lymph nodes

Background

Patients with melanoma and negative sentinel nodes (SNs) have varying outcomes, dependent on several prognostic factors. Considering all these factors in a prediction model might aid in identifying patients who could benefit from a personalized treatment strategy. The objective was to construct and validate a nomogram for recurrence and melanoma‐specific mortality (MSM) in patients with melanoma and negative SNs.

Methods

A total of 3220 patients with negative SNs were identified from a cohort of 4124 patients from four EORTC Melanoma Group centres who underwent sentinel lymph node biopsy. Prognostic factors for recurrence and MSM were studied with Cox regression analysis. Significant factors were incorporated in the models. Performance was assessed by discrimination (c‐index) and calibration in cross‐validation across the four centres. A nomogram was developed for graphical presentation.

Results

There were 3180 eligible patients. The final prediction model for recurrence and the calibrated model for MSM included three independent prognostic factors: ulceration, anatomical location and Breslow thickness. The c‐index was 0·74 for recurrence and 0·76 for the calibrated MSM model. Cross‐validation across the four centres showed reasonable model performance. A nomogram was developed based on these models. One‐third of the patients had a 5‐year recurrence probability of 8·2 per cent or less, and one‐third had a recurrence probability of 23·0 per cent or more.

Conclusion

A nomogram for predicting recurrence and MSM in patients with melanoma and negative SNs was constructed and validated. It could provide personalized estimates useful for tailoring surveillance strategies (reduce or increase intensity), and selection of patients for adjuvant therapy or clinical trials.

Could personalize care

Long non-coding RNA exchange during the oocyte-to-embryo transition in mice

The oocyte-to-embryo transition (OET) transforms a differentiated gamete into pluripotent blastomeres. The accompanying maternal-zygotic RNA exchange involves remodeling of the long non-coding RNA (lncRNA) pool. Here, we used next generation sequencing and de novo transcript assembly to define the core population of 1,600 lncRNAs expressed during the OET (lncRNAs). Relative to mRNAs, OET lncRNAs were less expressed and had shorter transcripts, mainly due to fewer exons and shorter 5′ terminal exons. Approximately half of OET lncRNA promoters originated in retrotransposons suggesting their recent emergence. Except for a small group of ubiquitous lncRNAs, maternal and zygotic lncRNAs formed two distinct populations. The bulk of maternal lncRNAs was degraded before the zygotic genome activation. Interestingly, maternal lncRNAs seemed to undergo cytoplasmic polyadenylation observed for dormant mRNAs. We also identified lncRNAs giving rise to trans-acting short interfering RNAs, which represent a novel lncRNA category. Altogether, we defined the core OET lncRNA transcriptome and characterized its remodeling during early development. Our results are consistent with the notion that rapidly evolving lncRNAs constitute signatures of cells-of-origin while a minority plays an active role in control of gene expression across OET. Our data presented here provide an excellent source for further OET lncRNA studies.

Widespread activation of antisense transcription of the host genome during herpes simplex virus 1 infection

Background

Herpesviruses can infect a wide range of animal species. Herpes simplex virus 1 (HSV-1) is one of the eight herpesviruses that can infect humans and is prevalent worldwide. Herpesviruses have evolved multiple ways to adapt the infected cells to their needs, but knowledge about these transcriptional and post-transcriptional modifications is sparse.

Results

Here, we show that HSV-1 induces the expression of about 1000 antisense transcripts from the human host cell genome. A subset of these is also activated by the closely related varicella zoster virus. Antisense transcripts originate either at gene promoters or within the gene body, and they show different susceptibility to the inhibition of early and immediate early viral gene expression. Overexpression of the major viral transcription factor ICP4 is sufficient to turn on a subset of antisense transcripts. Histone marks around transcription start sites of HSV-1-induced and constitutively transcribed antisense transcripts are highly similar, indicating that the genetic loci are already poised to transcribe these novel RNAs. Furthermore, an antisense transcript overlapping with the BBC3 gene (also known as PUMA) transcriptionally silences this potent inducer of apoptosis in cis.

Conclusions

We show for the first time that a virus induces widespread antisense transcription of the host cell genome. We provide evidence that HSV-1 uses this to downregulate a strong inducer of apoptosis. Our findings open new perspectives on global and specific alterations of host cell transcription by viruses.

Electronic supplementary material

The online version of this article (doi:10.1186/s13059-017-1329-5) contains supplementary material, which is available to authorized users.

RNA localization is a key determinant of neurite-enriched proteome

Protein subcellular localization is fundamental to the establishment of the body axis, cell migration, synaptic plasticity, and a vast range of other biological processes. Protein localization occurs through three mechanisms: protein transport, mRNA localization, and local translation. However, the relative contribution of each process to neuronal polarity remains unknown. Using neurons differentiated from mouse embryonic stem cells, we analyze protein and RNA expression and translation rates in isolated cell bodies and neurites genome-wide. We quantify 7323 proteins and the entire transcriptome, and identify hundreds of neurite-localized proteins and locally translated mRNAs. Our results demonstrate that mRNA localization is the primary mechanism for protein localization in neurites that may account for half of the neurite-localized proteome. Moreover, we identify multiple neurite-targeted non-coding RNAs and RNA-binding proteins with potential regulatory roles. These results provide further insight into the mechanisms underlying the establishment of neuronal polarity.

Subcellular localization of RNAs and proteins is important for polarized cells such as neurons. Here the authors differentiate mouse embryonic stem cells into neurons, and analyze the local transcriptome, proteome, and translated transcriptome in their cell bodies and neurites, providing a unique resource for future studies on neuronal polarity.

Long terminal repeats power evolution of genes and gene expression programs in mammalian oocytes and zygotes

Retrotransposons are "copy-and-paste" insertional mutagens that substantially contribute to mammalian genome content. Retrotransposons often carry long terminal repeats (LTRs) for retrovirus-like reverse transcription and integration into the genome. We report an extraordinary impact of a group of LTRs from the mammalian endogenous retrovirus-related ERVL retrotransposon class on gene expression in the germline and beyond. In mouse, we identified more than 800 LTRs from ORR1, MT, MT2, and MLT families, which resemble mobile gene-remodeling platforms that supply promoters and first exons. The LTR-mediated gene remodeling also extends to hamster, human, and bovine oocytes. The LTRs function in a stage-specific manner during the oocyte-to-embryo transition by activating transcription, altering protein-coding sequences, producing noncoding RNAs, and even supporting evolution of new protein-coding genes. These functions result, for example, in recycling processed pseudogenes into mRNAs or lncRNAs with regulatory roles. The functional potential of the studied LTRs is even higher, because we show that dormant LTR promoter activity can rescue loss of an essential upstream promoter. We also report a novel protein-coding gene evolution-D6Ertd527e-in which an MT LTR provided a promoter and the 5' exon with a functional start codon while the bulk of the protein-coding sequence evolved through a CAG repeat expansion. Altogether, ERVL LTRs provide molecular mechanisms for stochastically scanning, rewiring, and recycling genetic information on an extraordinary scale. ERVL LTRs thus offer means for a comprehensive survey of the genome's expression potential, tightly intertwining with gene expression and evolution in the germline.

Maternal SIN3A regulates reprogramming of gene expression during mouse preimplantation development

The oocyte-to-embryo transition entails genome activation and a dramatic reprogramming of gene expression that is required for continued development. Superimposed on genome activation and reprogramming is development of a transcriptionally repressive state at the level of chromatin structure. Inducing global histone hyperacetylation relieves this repression and histone deacetylases 1 and 2 (HDAC1 and HDAC2) are involved in establishing the repressive state. Because SIN3A is an HDAC1/2-containing complex, we investigated whether it is involved in reprogramming gene expression during the course of genome activation. We find that Sin3a mRNA is recruited during maturation and that inhibiting its recruitment not only inhibits development beyond the 2-cell stage but also compromises the fidelity of reprogramming gene expression. The SIN3A that is synthesized during oocyte maturation reaches a maximum level in the mid-1-cell embryo and is essentially absent by the mid-2-cell stage. Overexpressing SIN3A in 1-cell embryos has no obvious effect on pre- and postimplantation development. These results provide a mechanism by which reprogramming can occur using a maternally inherited transcription machinery, namely, recruitment of mRNAs encoding transcription factors and chromatin remodelers, such as SIN3A.

Sculpting the Transcriptome During the Oocyte-to-Embryo Transition in Mouse

In mouse, the oocyte-to-embryo transition entails converting a highly differentiated oocyte to totipotent blastomeres. This transition is driven by degradation of maternal mRNAs, which results in loss of oocyte identity, and reprogramming of gene expression during the course of zygotic gene activation, which occurs primarily during the two-cell stage and confers blastomere totipotency. Full-grown oocytes are transcriptionally quiescent and mRNAs are remarkably stable in oocytes due to the RNA-binding protein MSY2, which stabilizes mRNAs, and low activity of the 5' and 3' RNA degradation machinery. Oocyte maturation initiates a transition from mRNA stability to instability due to phosphorylation of MSY2, which makes mRNAs more susceptible to the RNA degradation machinery, and recruitment of dormant maternal mRNAs that encode for critical components of the 5' and 3' RNA degradation machinery. Small RNAs (miRNA, siRNA, and piRNA) play little, if any, role in mRNA degradation that occurs during maturation. Many mRNAs are totally degraded but a substantial fraction is only partially degraded, their degradation completed by the end of the two-cell stage. Genome activation initiates during the one-cell stage, is promiscuous, low level, and genome wide (and includes both inter- and intragenic regions) and produces transcripts that are inefficiently spliced and polyadenylated. The major wave of genome activation in two-cell embryos involves expression of thousands of new genes. This unique pattern of gene expression is the product of maternal mRNAs recruited during maturation that encode for transcription factors and chromatin remodelers, as well as dramatic changes in chromatin structure due to incorporation of histone variants and modified histones.

The first murine zygotic transcription is promiscuous and uncoupled from splicing and 3' processing

Initiation of zygotic transcription in mammals is poorly understood. In mice, zygotic transcription is first detected shortly after pronucleus formation in 1-cell embryos, but the identity of the transcribed loci and mechanisms regulating their expression are not known. Using total RNA-Seq, we have found that transcription in 1-cell embryos is highly promiscuous, such that intergenic regions are extensively expressed and thousands of genes are transcribed at comparably low levels. Striking is that transcription can occur in the absence of defined core-promoter elements. Furthermore, accumulation of translatable zygotic mRNAs is minimal in 1-cell embryos because of inefficient splicing and 3' processing of nascent transcripts. These findings provide novel insights into regulation of gene expression in 1-cell mouse embryos that may confer a protective mechanism against precocious gene expression that is the product of a relaxed chromatin structure present in 1-cell embryos. The results also suggest that the first zygotic transcription itself is an active component of chromatin remodeling in 1-cell embryos.

Genomation: a toolkit to summarize, annotate and visualize genomic intervals

Biological insights can be obtained through computational integration of genomics data sets consisting of diverse types of information. The integration is often hampered by a large variety of existing file formats, often containing similar information, and the necessity to use complicated tools to achieve the desired results. We have built an R package, genomation, to expedite the extraction of biological information from high throughput data. The package works with a variety of genomic interval file types and enables easy summarization and annotation of high throughput data sets with given genomic annotations.

AVAILABILITY AND IMPLEMENTATION

The software is currently distributed under MIT artistic license and freely available at http://bioinformatics.mdc-berlin.de/genomation, and through the Bioconductor framework.

Archaeal aminoacyl-tRNA synthetases interact with the ribosome to recycle tRNAs

Aminoacyl-tRNA synthetases (aaRS) are essential enzymes catalyzing the formation of aminoacyl-tRNAs, the immediate precursors for encoded peptides in ribosomal protein synthesis. Previous studies have suggested a link between tRNA aminoacylation and high-molecular-weight cellular complexes such as the cytoskeleton or ribosomes. However, the structural basis of these interactions and potential mechanistic implications are not well understood. To biochemically characterize these interactions we have used a system of two interacting archaeal aaRSs: an atypical methanogenic-type seryl-tRNA synthetase and an archaeal ArgRS. More specifically, we have shown by thermophoresis and surface plasmon resonance that these two aaRSs bind to the large ribosomal subunit with micromolar affinities. We have identified the L7/L12 stalk and the proteins located near the stalk base as the main sites for aaRS binding. Finally, we have performed a bioinformatics analysis of synonymous codons in the Methanothermobacter thermautotrophicus genome that supports a mechanism in which the deacylated tRNAs may be recharged by aaRSs bound to the ribosome and reused at the next occurrence of a codon encoding the same amino acid. These results suggest a mechanism of tRNA recycling in which aaRSs associate with the L7/L12 stalk region to recapture the tRNAs released from the preceding ribosome in polysomes.

A cohesin-independent role for NIPBL at promoters provides insights in CdLS

The cohesin complex is crucial for chromosome segregation during mitosis and has recently also been implicated in transcriptional regulation and chromatin architecture. The NIPBL protein is required for the loading of cohesin onto chromatin, but how and where cohesin is loaded in vertebrate cells is unclear. Heterozygous mutations of NIPBL were found in 50% of the cases of Cornelia de Lange Syndrome (CdLS), a human developmental syndrome with a complex phenotype. However, no defects in the mitotic function of cohesin have been observed so far and the links between NIPBL mutations and the observed developmental defects are unclear. We show that NIPBL binds to chromatin in somatic cells with a different timing than cohesin. Further, we observe that high-affinity NIPBL binding sites localize to different regions than cohesin and almost exclusively to the promoters of active genes. NIPBL or cohesin knockdown reduce transcription of these genes differently, suggesting a cohesin-independent role of NIPBL for transcription. Motif analysis and comparison to published data show that NIPBL co-localizes with a specific set of other transcription factors. In cells derived from CdLS patients NIPBL binding levels are reduced and several of the NIPBL-bound genes have previously been observed to be mis-expressed in CdLS. In summary, our observations indicate that NIPBL mutations might cause developmental defects in different ways. First, defects of NIPBL might lead to cohesin-loading defects and thereby alter gene expression and second, NIPBL deficiency might affect genes directly via its role at the respective promoters.

The male germ cell gene regulator CTCFL is functionally different from CTCF and binds CTCF-like consensus sites in a nucleosome composition-dependent manner

Background

CTCF is a highly conserved and essential zinc finger protein expressed in virtually all cell types. In conjunction with cohesin, it organizes chromatin into loops, thereby regulating gene expression and epigenetic events. The function of CTCFL or BORIS, the testis-specific paralog of CTCF, is less clear.

Results

Using immunohistochemistry on testis sections and fluorescence-based microscopy on intact live seminiferous tubules, we show that CTCFL is only transiently present during spermatogenesis, prior to the onset of meiosis, when the protein co-localizes in nuclei with ubiquitously expressed CTCF. CTCFL distribution overlaps completely with that of Stra8, a retinoic acid-inducible protein essential for the propagation of meiosis. We find that absence of CTCFL in mice causes sub-fertility because of a partially penetrant testicular atrophy. CTCFL deficiency affects the expression of a number of testis-specific genes, including Gal3st1 and Prss50. Combined, these data indicate that CTCFL has a unique role in spermatogenesis. Genome-wide RNA expression studies in ES cells expressing a V5- and GFP-tagged form of CTCFL show that genes that are downregulated in CTCFL-deficient testis are upregulated in ES cells. These data indicate that CTCFL is a male germ cell gene regulator. Furthermore, genome-wide DNA-binding analysis shows that CTCFL binds a consensus sequence that is very similar to that of CTCF. However, only ~3,700 out of the ~5,700 CTCFL- and ~31,000 CTCF-binding sites overlap. CTCFL binds promoters with loosely assembled nucleosomes, whereas CTCF favors consensus sites surrounded by phased nucleosomes. Finally, an ES cell-based rescue assay shows that CTCFL is functionally different from CTCF.

Conclusions

Our data suggest that nucleosome composition specifies the genome-wide binding of CTCFL and CTCF. We propose that the transient expression of CTCFL in spermatogonia and preleptotene spermatocytes serves to occupy a subset of promoters and maintain the expression of male germ cell genes.

Using cloud computing infrastructure with CloudBioLinux, CloudMan, and Galaxy

Cloud computing has revolutionized availability and access to computing and storage resources, making it possible to provision a large computational infrastructure with only a few clicks in a Web browser. However, those resources are typically provided in the form of low-level infrastructure components that need to be procured and configured before use. In this unit, we demonstrate how to utilize cloud computing resources to perform open-ended bioinformatic analyses, with fully automated management of the underlying cloud infrastructure. By combining three projects, CloudBioLinux, CloudMan, and Galaxy, into a cohesive unit, we have enabled researchers to gain access to more than 100 preconfigured bioinformatics tools and gigabytes of reference genomes on top of the flexible cloud computing infrastructure. The protocol demonstrates how to set up the available infrastructure and how to use the tools via a graphical desktop interface, a parallel command-line interface, and the Web-based Galaxy interface.

Prediction of interacting protein residues using sequence and structure data

Identifying hotspots responsible for protein interactions with other macromolecules or drugs provides insight into functional aspects of the protein network, and is a pivotal task in systems biology and drug discovery. Here, we present the protocol for the application of a machine-learning method - Random Forest - to prediction of interacting residues in proteins, based on either the structural parameters or the primary sequence alone.

Characterisation of deep arterio-venous anastomoses within monochorionic placentae by vascular casting

OBJECTIVE

To characterise arterio-venous anastomoses (AVA) in monochorionic (MC) placentae and determine (i) whether shared cotyledons lie beneath the co-termination of an artery from one twin and a vein to the contralateral twin and (ii) whether all AVA can be detected by visual inspection of the chorionic plate.

METHODS

Vascular casts were made of 15 MC placentae. The number of typical AVAs suspected visually before digestion was compared with the number of AVAs identified after acid digestion.

RESULTS

Thirty-three of 67 (49%) suspected typical AVAs were confirmed as typical after casting. There were five false positives and no false negatives. The remainder were classified as atypical AVAs, found in > or =90% of MC placentae. Type I (small vascular connections between two apparently normal cotyledons not seen before casting) and Type II (shared cotyledons arising within larger apparently normal cotyledons) atypical AVAs were found in 53% and 73% of placentae, respectively.

CONCLUSIONS

Only half the shared cotyledons in MC placentae are characterised by co-termination of an artery and vein on the chorionic plate. We report the existence of deep anastomoses beneath the chorionic plate that cannot be visualised by chorionic plate inspection. These findings have implications for laser treatment of twin-twin transfusion syndrome.