A deep learning framework for predicting human essential genes from population and functional genomic data

Being able to predict essential genes intolerant to loss-of-function (LOF) mutations can dramatically improve our ability to identify genes associated with genetic disorders. Numerous computational methods have recently been developed to predict human essential genes from population genomic data; however, the existing methods have limited power in pinpointing short essential genes due to the sparsity of polymorphisms in the human genome. Here we present an evolution-based deep learning model, DeepLOF, which integrates population and functional genomic data to improve gene essentiality prediction. Compared to previous methods, DeepLOF shows unmatched performance in predicting ClinGen haploinsufficient genes, mouse essential genes, and essential genes in human cell lines. Furthermore, DeepLOF discovers 109 potentially essential genes that are too short to be identified by previous methods. Altogether, DeepLOF is a powerful computational method to aid in the discovery of essential genes.

clipplotr - a comparative visualisation and analysis tool for CLIP data

CLIP technologies are now widely used to study RNA-protein interactions and many datasets are now publicly available. An important first step in CLIP data exploration is the visual inspection and assessment of processed genomic data on selected genes or regions and performing comparisons: either across conditions within a particular project, or incorporating publicly available data. However, the output files produced by data processing pipelines or preprocessed files available to download from data repositories are often not suitable for direct comparison and usually need further processing. Furthermore, to derive biological insight it is usually necessary to visualise CLIP signal alongside other data such as annotations, or orthogonal functional genomic data (e.g. RNA-seq). We have developed a simple, but powerful, command-line tool: clipplotr, which facilitates these visual comparative and integrative analyses with normalisation and smoothing options for CLIP data and the ability to show these alongside reference annotation tracks and functional genomic data. These data can be supplied as input to clipplotr in a range of file formats, which will output a publication quality figure. It is written in R and can both run on a laptop computer independently, or be integrated into computational workflows on a high-performance cluster. Releases, source code and documentation are freely available at: https://github.com/ulelab/clipplotr.

Selection against admixture and gene regulatory divergence in a long-term primate field study

Admixture has profoundly influenced evolution across the tree of life, including in humans and other primates. However, we have limited insight into the genetic and phenotypic consequences of admixture in primates, especially during its key early stages. Here, we address this gap by combining 50 years of field observations with population and functional genomic data from yellow (Papio cynocephalus) and anubis (P. anubis) baboons in Kenya, in a longitudinally studied population that has experienced both historical and recent admixture. We use whole-genome sequencing to characterize the extent of the hybrid zone, estimate local ancestry for 442 known individuals, and predict the landscape of introgression across the genome. Despite no major fitness costs to hybrids, we identify signatures of selection against introgression that are strikingly similar to those described for archaic hominins. These signatures are strongest near loci with large ancestry effects on gene expression, supporting the importance of gene regulation in primate evolution and the idea that selection targeted large regulatory effects following archaic hominin admixture. Our results show that genomic data and field observations of hybrids are important and mutually informative. They therefore demonstrate the value of other primates as living models for phenomena that we cannot observe in our own lineage.

Evidences for functional trans-acting eRNA-promoter R-loops at Alu sequences

Enhancers modulate gene expression by interacting with promoters. Models of enhancer-promoter interactions (EPIs) in the literature involve the activity of many components, including transcription factors and nucleic acid. However, the role that sequence similarity plays in EPIs, remains largely unexplored. Herein, we report that Alu-derived sequences dominate sequence similarity between enhancers and promoters. After rejecting the alternative DNA:DNA and DNA:RNA triplex models, we proposed that enhancer-associated RNAs, or eRNAs, may directly contact their targeted promoters by forming trans-acting R-loops at those Alu sequences. We showed how the characteristic distribution of functional genomic data, such as RNA-DNA proximate ligation reads, binding of transcription factors, and RNA-binding proteins, align with the Alu sequences of EPIs. We also showed that these aligned Alu sequences may be subject to the constraint of coevolution, further implying the functional significance of these R-loop hybrids. Finally, our results showed that eRNA and Alu elements associate in a manner previously unrecognized in the EPIs and the evolution of gene regulation networks in mammals.

PeakMatcher facilitates updated Aedes aegypti embryonic cis-regulatory element map

Background The Aedes aegypti mosquito is a threat to human health across the globe. The A. aegypti genome was recently re-sequenced and re-assembled. Due to a combination of long-read PacBio and Hi-C sequencing, the AaegL5 assembly is chromosome complete and significantly improves the assembly in key areas such as the M/m sex-determining locus. Release of the updated genome assembly has precipitated the need to reprocess historical functional genomic data sets, including cis-regulatory element (CRE) maps that had previously been generated for A. aegypti. Results We re-processed and re-analyzed the A. aegypti whole embryo FAIRE seq data to create an updated embryonic CRE map for the AaegL5 genome. We validated that the new CRE map recapitulates key features of the original AaegL3 CRE map. Further, we built on the improved assembly in the M/m locus to analyze overlaps of open chromatin regions with genes. To support the validation, we created a new method (PeakMatcher) for matching peaks from the same experimental data set across genome assemblies. Conclusion Use of PeakMatcher software, which is available publicly under an open-source license, facilitated the release of an updated and validated CRE map, which is available through the NIH GEO. These findings demonstrate that PeakMatcher software will be a useful resource for validation and transferring of previous annotations to updated genome assemblies.

Chinese Glioma Genome Atlas (CGGA): A Comprehensive Resource with Functional Genomic Data for Chinese Glioma Patients

Gliomas are the most common and malignant intracranial tumours in adults. Recent studies have shown that functional genomics greatly aids in the understanding of the pathophysiology and therapy of glioma. However, comprehensive genomic data and analysis platforms are relatively limited. In this study, we developed the Chinese Glioma Genome Atlas (CGGA, http://www.cgga.org.cn), a user-friendly data portal for storage and interactive exploration of multi-dimensional functional genomic data that includes nearly 2,000 primary and recurrent glioma samples from Chinese cohorts. CGGA currently provides access to whole-exome sequencing (286 samples), messenger RNA sequencing (1,018 samples) and microarray (301 samples), DNA methylation microarray (159 samples), and microRNA microarray (198 samples) data, as well as detailed clinical data (e.g., WHO grade, histological type, critical molecular genetic information, age, sex, chemoradiotherapy status and survival data). In addition, we developed an analysis tool to allow users to browse mutational, mRNA/microRNA expression, and DNA methylation profiles and perform survival and correlation analyses of specific glioma subtypes. CGGA greatly reduces the barriers between complex functional genomic data and glioma researchers who seek rapid, intuitive, and high-quality access to data resources and enables researchers to use these immeasurable data sources for biological research and clinical application. Importantly, the free provision of data will allow researchers to quickly generate and provide data to the research community.

YeasTSS: An Integrative Web Database of Yeast Transcription Start Sites

The transcription initiation landscape of eukaryotic genes is complex and highly dynamic. In eukaryotes, genes can generate multiple transcript variants that differ in 5’ boundaries due to usages of alternative transcription start sites (TSSs), and the abundance of transcript isoforms are highly variable. Due to a large number and complexity of the TSSs, it is not feasible to depict details of transcript initiation landscape of all genes using text-format genome annotation files. Therefore, it is necessary to provide data visualization of TSSs to represent quantitative TSS maps and the core promoters. In addition, the selection and activity of TSSs are influenced by various factors, such as transcription factors, chromatin remodeling, and histone modifications. Thus, integration and visualization of functional genomic data related to these features could provide a better understanding of the gene promoter architecture and regulatory mechanism of transcription initiation. Yeast species play important roles for the research and human society, yet no database provides visualization and integration of functional genomic data in yeast. Here, we generated quantitative TSS maps for twelve important yeast species, inferred their core promoters, and built a public database, YeasTSS (www.yeastss.org). YeasTSS was designed as a central portal for visualization and integration of the TSS maps, core promoters and functional genomic data related to transcription initiation in yeast. YeasTSS is expected to benefit the research community and public education for improving genome annotation, studies of promoter structure, regulated control of transcription initiation and inferring gene regulatory network.

YeasTSS: an integrative web database of yeast transcription start sites

The transcription initiation landscape of eukaryotic genes is complex and highly dynamic. In eukaryotes, genes can generate multiple transcript variants that differ in 5′ boundaries due to usages of alternative transcription start sites (TSSs), and the abundance of transcript isoforms are highly variable. Due to a large number and complexity of the TSSs, it is not feasible to depict details of transcript initiation landscape of all genes using text-format genome annotation files. Therefore, it is necessary to provide data visualization of TSSs to represent quantitative TSS maps and the core promoters (CPs). In addition, the selection and activity of TSSs are influenced by various factors, such as transcription factors, chromatin remodeling and histone modifications. Thus, integration and visualization of functional genomic data related to these features could provide a better understanding of the gene promoter architecture and regulatory mechanism of transcription initiation. Yeast species play important roles for the research and human society, yet no database provides visualization and integration of functional genomic data in yeast. Here, we generated quantitative TSS maps for 12 important yeast species, inferred their CPs and built a public database, YeasTSS (www.yeastss.org). YeasTSS was designed as a central portal for visualization and integration of the TSS maps, CPs and functional genomic data related to transcription initiation in yeast. YeasTSS is expected to benefit the research community and public education for improving genome annotation, studies of promoter structure, regulated control of transcription initiation and inferring gene regulatory network.