The African Society of Human Genetics successfully launches global data science workshops.

To accelerate the impact of African genomics on human health, data science skills and awareness of Africa's rich genetic diversity must be strengthened globally. We describe the first African genomics data science workshop, implemented by the African Society of Human Genetics (AfSHG) and international partners, providing a framework for future workshops.

Performance assessment of computational tools to detect microsatellite instability.

Microsatellite instability (MSI) is a phenomenon seen in several cancer types, which can be used as a biomarker to help guide immune checkpoint inhibitor treatment. To facilitate this, researchers have developed computational tools to categorize samples as having high microsatellite instability, or as being microsatellite stable using next-generation sequencing data. Most of these tools were published with unclear scope and usage, and they have yet to be independently benchmarked. To address these issues, we assessed the performance of eight leading MSI tools across several unique datasets that encompass a wide variety of sequencing methods. While we were able to replicate the original findings of each tool on whole exome sequencing data, most tools had worse receiver operating characteristic and precision-recall area under the curve values on whole genome sequencing data. We also found that they lacked agreement with one another and with commercial MSI software on gene panel data, and that optimal threshold cut-offs vary by sequencing type. Lastly, we tested tools made specifically for RNA sequencing data and found they were outperformed by tools designed for use with DNA sequencing data. Out of all, two tools (MSIsensor2, MANTIS) performed well across nearly all datasets, but when all datasets were combined, their precision decreased. Our results caution that MSI tools can have much lower performance on datasets other than those on which they were originally evaluated, and in the case of RNA sequencing tools, can even perform poorly on the type of data for which they were created.

Regulatory landscape enrichment analysis (RLEA): a computational toolkit for non-coding variant enrichment and cell type prioritization.

BACKGROUND: As genomic studies continue to implicate non-coding sequences in disease, testing the roles of these variants requires insights into the cell type(s) in which they are likely to be mediating their effects. Prior methods for associating non-coding variants with cell types have involved approaches using linkage disequilibrium or ontological associations, incurring significant processing requirements. GaiaAssociation is a freely available, open-source software that enables thousands of genomic loci implicated in a phenotype to be tested for enrichment at regulatory loci of multiple cell types in minutes, permitting insights into the cell type(s) mediating the studied phenotype. RESULTS: In this work, we present Regulatory Landscape Enrichment Analysis (RLEA) by GaiaAssociation and demonstrate its capability to test the enrichment of 12,133 variants across the cis-regulatory regions of 44 cell types. This analysis was completed in 134.0 ± 2.3 s, highlighting the efficient processing provided by GaiaAssociation. The intuitive interface requires only four inputs, offers a collection of customizable functions, and visualizes variant enrichment in cell-type regulatory regions through a heatmap matrix. GaiaAssociation is available on PyPi for download as a command line tool or Python package and the source code can also be installed from GitHub at https://github.com/GreallyLab/gaiaAssociation . CONCLUSIONS: GaiaAssociation is a novel package that provides an intuitive and efficient resource to understand the enrichment of non-coding variants across the cis-regulatory regions of different cells, empowering studies seeking to identify disease-mediating cell types.

Random-effects meta-analysis of effect sizes as a unified framework for gene set analysis.

Gene set analysis (GSA) remains a common step in genome-scale studies because it can reveal insights that are not apparent from results obtained for individual genes. Many different computational tools are applied for GSA, which may be sensitive to different types of signals; however, most methods implicitly test whether there are differences in the distribution of the effect of some experimental condition between genes in gene sets of interest. We have developed a unifying framework for GSA that first fits effect size distributions, and then tests for differences in these distributions between gene sets. These differences can be in the proportions of genes that are perturbed or in the sign or size of the effects. Inspired by statistical meta-analysis, we take into account the uncertainty in effect size estimates by reducing the influence of genes with greater uncertainty on the estimation of distribution parameters. We demonstrate, using simulation and by application to real data, that this approach provides significant gains in performance over existing methods. Furthermore, the statistical tests carried out are defined in terms of effect sizes, rather than the results of prior statistical tests measuring these changes, which leads to improved interpretability and greater robustness to variation in sample sizes.

No evidence that HLA genotype influences the driver mutations that occur in cancer patients.

The major histocompatibility (MHC) molecules are capable of presenting neoantigens resulting from somatic mutations on cell surfaces, potentially directing immune responses against cancer. This led to the hypothesis that cancer driver mutations may occur in gaps in the capacity to present neoantigens that are dependent on MHC genotype. If this is correct, it has important implications for understanding oncogenesis and may help to predict driver mutations based on genotype data. In support of this hypothesis, it has been reported that driver mutations that occur frequently tend to be poorly presented by common MHC alleles and that the capacity of a patient's MHC alleles to present the resulting neoantigens is predictive of the driver mutations that are observed in their tumor. Here we show that these reports of a strong relationship between driver mutation occurrence and patient MHC alleles are a consequence of unjustified statistical assumptions. Our reanalysis of the data provides no evidence of an effect of MHC genotype on the oncogenic mutation landscape.

Bias and inconsistency in the estimation of tumour mutation burden.

BACKGROUND: Tumour mutation burden (TMB), defined as the number of somatic mutations per megabase within the sequenced region in the tumour sample, has been used as a biomarker for predicting response to immune therapy. Several studies have been conducted to assess the utility of TMB for various cancer types; however, methods to measure TMB have not been adequately evaluated. In this study, we identified two sources of bias in current methods to calculate TMB. METHODS: We used simulated data to quantify the two sources of bias and their effect on TMB calculation, we down-sampled sequencing reads from exome sequencing datasets from TCGA to evaluate the consistency in TMB estimation across different sequencing depths. We analyzed data from ten cancer cohorts to investigate the relationship between inferred TMB and sequencing depth. RESULTS: We found that TMB, estimated by counting the number of somatic mutations above a threshold frequency (typically 0.05), is not robust to sequencing depth. Furthermore, we show that, because only mutations with an observed frequency greater than the threshold are considered, the observed mutant allele frequency provides a biased estimate of the true frequency. This can result in substantial over-estimation of the TMB, when the cancer sample includes a large number of somatic mutations at low frequencies, and exacerbates the lack of robustness of TMB to variation in sequencing depth and tumour purity. CONCLUSION: Our results demonstrate that care needs to be taken in the estimation of TMB to ensure that results are unbiased and consistent across studies and we suggest that accurate and robust estimation of TMB could be achieved using statistical models that estimate the full mutant allele frequency spectrum.

Selection Shapes Synonymous Stop Codon Use in Mammals.

Phylogenetic models of the evolution of protein-coding sequences can provide insights into the selection pressures that have shaped them. In the application of these models synonymous nucleotide substitutions, which do not alter the encoded amino acid, are often assumed to have limited functional consequences and used as a proxy for the neutral rate of evolution. The ratio of nonsynonymous to synonymous substitution rates is then used to categorize the selective regime that applies to the protein (e.g., purifying selection, neutral evolution, diversifying selection). Here, we extend the Muse and Gaut model of codon evolution to explore the extent of purifying selection acting on substitutions between synonymous stop codons. Using a large collection of coding sequence alignments, we estimate that a high proportion (approximately 57%) of mammalian genes are affected by selection acting on stop codon preference. This proportion varies substantially by codon, with UGA stop codons far more likely to be conserved. Genes with evidence of selection acting on synonymous stop codons have distinctive characteristics, compared to unconserved genes with the same stop codon, including longer [Formula: see text] untranslated regions (UTRs) and shorter mRNA half-life. The coding regions of these genes are also much more likely to be under strong purifying selection pressure. Our results suggest that the preference for UGA stop codons found in many multicellular eukaryotes is selective rather than mutational in origin.

Inference of Candidate Germline Mutator Loci in Humans from Genome-Wide Haplotype Data.

The rate of germline mutation varies widely between species but little is known about the extent of variation in the germline mutation rate between individuals of the same species. Here we demonstrate that an allele that increases the rate of germline mutation can result in a distinctive signature in the genomic region linked to the affected locus, characterized by a number of haplotypes with a locally high proportion of derived alleles, against a background of haplotypes carrying a typical proportion of derived alleles. We searched for this signature in human haplotype data from phase 3 of the 1000 Genomes Project and report a number of candidate mutator loci, several of which are located close to or within genes involved in DNA repair or the DNA damage response. To investigate whether mutator alleles remained active at any of these loci, we used de novo mutation counts from human parent-offspring trios in the 1000 Genomes and Genome of the Netherlands cohorts, looking for an elevated number of de novo mutations in the offspring of parents carrying a candidate mutator haplotype at each of these loci. We found some support for two of the candidate loci, including one locus just upstream of the BRSK2 gene, which is expressed in the testis and has been reported to be involved in the response to DNA damage.

A Database of Human Immune Receptor Alleles Recovered from Population Sequencing Data.

High-throughput sequencing data from TCRs and Igs can provide valuable insights into the adaptive immune response, but bioinformatics pipelines for analysis of these data are constrained by the availability of accurate and comprehensive repositories of TCR and Ig alleles. We have created an analytical pipeline to recover immune receptor alleles from genome sequencing data. Applying this pipeline to data from the 1000 Genomes Project we have created Lym1K, a collection of immune receptor alleles that combines known, well-supported alleles with novel alleles found in the 1000 Genomes Project data. We show that Lym1K leads to a significant improvement in the alignment of short read sequences from immune receptors and that the addition of novel alleles discovered from genome sequence data are likely to be particularly significant for comprehensive analysis of populations that are not currently well represented in existing repositories of immune alleles.

LymAnalyzer: a tool for comprehensive analysis of next generation sequencing data of T cell receptors and immunoglobulins.

The adaptive immune system includes populations of B and T cells capable of binding foreign epitopes via antigen specific receptors, called immunoglobulin (IG) for B cells and the T cell receptor (TCR) for T cells. In order to provide protection from a wide range of pathogens, these cells display highly diverse repertoires of IGs and TCRs. This is achieved through combinatorial rearrangement of multiple gene segments in addition, for B cells, to somatic hypermutation. Deep sequencing technologies have revolutionized analysis of the diversity of these repertoires; however, accurate TCR/IG diversity profiling requires specialist bioinformatics tools. Here we present LymAnalzyer, a software package that significantly improves the completeness and accuracy of TCR/IG profiling from deep sequence data and includes procedures to identify novel alleles of gene segments. On real and simulated data sets LymAnalyzer produces highly accurate and complete results. Although, to date we have applied it to TCR/IG data from human and mouse, it can be applied to data from any species for which an appropriate database of reference genes is available. Implemented in Java, it includes both a command line version and a graphical user interface and is freely available at https://sourceforge.net/projects/lymanalyzer/.

A genomic portrait of haplotype diversity and signatures of selection in indigenous southern African populations.

We report a study of genome-wide, dense SNP (∼ 900K) and copy number polymorphism data of indigenous southern Africans. We demonstrate the genetic contribution to southern and eastern African populations, which involved admixture between indigenous San, Niger-Congo-speaking and populations of Eurasian ancestry. This finding illustrates the need to account for stratification in genome-wide association studies, and that admixture mapping would likely be a successful approach in these populations. We developed a strategy to detect the signature of selection prior to and following putative admixture events. Several genomic regions show an unusual excess of Niger-Kordofanian, and unusual deficiency of both San and Eurasian ancestry, which were considered the footprints of selection after population admixture. Several SNPs with strong allele frequency differences were observed predominantly between the admixed indigenous southern African populations, and their ancestral Eurasian populations. Interestingly, many candidate genes, which were identified within the genomic regions showing signals for selection, were associated with southern African-specific high-risk, mostly communicable diseases, such as malaria, influenza, tuberculosis, and human immunodeficiency virus/AIDs. This observation suggests a potentially important role that these genes might have played in adapting to the environment. Additionally, our analyses of haplotype structure, linkage disequilibrium, recombination, copy number variation and genome-wide admixture highlight, and support the unique position of San relative to both African and non-African populations. This study contributes to a better understanding of population ancestry and selection in south-eastern African populations; and the data and results obtained will support research into the genetic contributions to infectious as well as non-communicable diseases in the region.

Intron Length Coevolution across Mammalian Genomes.

Although they do not contribute directly to the proteome, introns frequently contain regulatory elements and can extend the protein coding potential of the genome through alternative splicing. For some genes, the contribution of introns to the time required for transcription can also be functionally significant. We have previously shown that intron length in genes associated with developmental patterning is often highly conserved. In general, sets of genes that require precise coordination in the timing of their expression may be sensitive to changes in transcript length. A prediction of this hypothesis is that evolutionary changes in intron length, when they occur, may be correlated between sets of coordinately expressed genes. To test this hypothesis, we analyzed intron length coevolution in alignments from nine eutherian mammals. Overall, genes that belong to the same protein complex or that are coexpressed were significantly more likely to show evidence of intron length coevolution than matched, randomly sampled genes. Individually, protein complexes involved in the cell cycle showed the strongest evidence of coevolution of intron lengths and clusters of coexpressed genes enriched for cell cycle genes also showed significant evidence of intron length coevolution. Our results reveal a novel aspect of gene coevolution and provide a means to identify genes, protein complexes and biological processes that may be particularly sensitive to changes in transcriptional dynamics.

Differential DNA methylation patterns of homeobox genes in proximal and distal colon epithelial cells.

Region and cell-type specific differences in the molecular make up of colon epithelial cells have been reported. Those differences may underlie the region-specific characteristics of common colon epithelial diseases such as colorectal cancer and inflammatory bowel disease. DNA methylation is a cell-type specific epigenetic mark, essential for transcriptional regulation, silencing of repetitive DNA and genomic imprinting. Little is known about any region-specific variations in methylation patterns in human colon epithelial cells. Using purified epithelial cells and whole biopsies (n= 19) from human subjects, we generated epigenome-wide DNA methylation data (using the HELP-tagging assay), comparing the methylation signatures of the proximal and distal colon. We identified a total of 125 differentially methylated sites (DMS) mapping to transcription start sites of protein-coding genes, most notably several members of the homeobox (HOX) family of genes. Patterns of differential methylation were validated with MassArray EpiTYPER. We also examined DNA methylation in whole biopsies, applying a computational technique to deconvolve variation in methylation within cell types and variation in cell-type composition across biopsies. Including inferred epithelial proportions as a covariate in differential methylation analysis applied to the whole biopsies resulted in greater overlap with the results obtained from purified epithelial cells compared with when the covariate was not included. Results obtained from both approaches highlight region-specific methylation patterns of HOX genes in colonic epithelium. Regional variation in methylation patterns has implications for the study of diseases that exhibit regional expression patterns in the human colon, such as inflammatory bowel disease and colorectal cancer.

The shared genomic architecture of human nucleolar organizer regions.

The short arms of the five acrocentric human chromosomes harbor sequences that direct the assembly and function of the nucleolus, one of the key functional domains of the nucleus, yet they are absent from the current human genome assembly. Here we describe the genomic architecture of these human nucleolar organizers. Sequences distal and proximal to ribosomal gene arrays are conserved among the acrocentric chromosomes, suggesting they are sites of frequent recombination. Although previously believed to be heterochromatic, characterization of these two flanking regions reveals that they share a complex genomic architecture similar to other euchromatic regions of the genome, but they have distinct genomic characteristics. Proximal sequences are almost entirely segmentally duplicated, similar to the regions bordering centromeres. In contrast, the distal sequence is predominantly unique to the acrocentric short arms and is dominated by a very large inverted repeat. We show that the distal element is localized to the periphery of the nucleolus, where it appears to anchor the ribosomal gene repeats. This, combined with its complex chromatin structure and transcriptional activity, suggests that this region is involved in nucleolar organization. Our results provide a platform for investigating the role of NORs in nucleolar formation and function, and open the door for determining the role of these regions in the well-known empirical association of nucleoli with pathology.

Promiscuous mRNA splicing under the control of AIRE in medullary thymic epithelial cells.

MOTIVATION: The expression of tissue-restricted antigens (TRAs) in the thymus is required to ensure efficient negative selection of potentially auto-reactive T lymphocytes and avoid autoimmune disease. This promiscuous expression is under the control of the autoimmune regulator (AIRE), a transcription factor expressed in medullary thymic epithelial cells (mTECs). Tissue-specific alternative splicing may also produce TRAs but the extent to which splice isoforms that are restricted to specific tissues are expressed in mTECs is yet to be investigated. RESULTS: We reanalyzed microarray and RNA-Seq datasets from mouse mTECs and other epithelial and non-epithelial cell types and found that the diversity of splice isoforms in mTECs was greater than in any of the other cell types or tissues studied. We identified tissue-specific isoforms from a panel of mouse tissues and found several examples of such isoforms that are expressed in mTECs. The number of isoforms with restricted expression found in mTECs was significantly higher than for comparable cell types. Furthermore, we found evidence that AIRE influences the increased splicing diversity observed in mTECs as the genes for which tissue restricted isoforms are produced in mTECs were significantly more likely than other genes to be differentially spliced between AIRE knock-out and wild-type samples. Our results suggest that developing T lymphocytes are exposed to diverse tissue-restricted splice isoforms in the thymus and that AIRE has a direct or indirect role in this process, representing a novel aspect of its role in the maintenance of immune self-tolerance. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Seq-ing improved gene expression estimates from microarrays using machine learning.

BACKGROUND: Quantifying gene expression by RNA-Seq has several advantages over microarrays, including greater dynamic range and gene expression estimates on an absolute, rather than a relative scale. Nevertheless, microarrays remain in widespread use, demonstrated by the ever-growing numbers of samples deposited in public repositories. RESULTS: We propose a novel approach to microarray analysis that attains many of the advantages of RNA-Seq. This method, called Machine Learning of Transcript Expression (MaLTE), leverages samples for which both microarray and RNA-Seq data are available, using a Random Forest to learn the relationship between the fluorescence intensity of sets of microarray probes and RNA-Seq transcript expression estimates. We trained MaLTE on data from the Genotype-Tissue Expression (GTEx) project, consisting of Affymetrix gene arrays and RNA-Seq from over 700 samples across a broad range of human tissues. CONCLUSION: This approach can be used to accurately estimate absolute expression levels from microarray data, at both gene and transcript level, which has not previously been possible. This methodology will facilitate re-analysis of archived microarray data and broaden the utility of the vast quantities of data still being generated.

Alterations in hepatic miRNA expression during negative energy balance in postpartum dairy cattle.

BACKGROUND: Negative energy balance (NEB), an altered metabolic state, occurs in early postpartum dairy cattle when energy demands to support lactation exceed energy intake. During NEB the liver undergoes oxidative stress and increased breakdown of fatty acids accompanied by changes in gene expression. It is now known that micro RNAs (miRNA) can have a role in mediating such alterations in gene expression through repression or degradation of target mRNAs. miRNA expression is known to be altered by metabolism and environmental factors and miRNAs are implicated in expression modulation of metabolism related genes. RESULTS: miRNA expression was profiled in the liver of moderate yielding dairy cattle under severe NEB (SNEB) and mild NEB (MNEB) using the Affymetrix Gene Chip miRNA_2.0 array with 679 probe sets for Bos-taurus miRNAs. Ten miRNAs were found to be differentially expressed using the 'samr' statistical package (delta = 0.6) at a q-value FDR of < 12%. Five miRNAs including miR-17-5p, miR-31, miR-140, miR-1281 and miR-2885 were validated using RT-qPCR, to be up-regulated under SNEB. Liver diseases associated with these miRNAs include non-alcoholic fatty liver (NAFLD) and hepatocellular carcinoma (HCC). miR-140 and miR-17-5p are known to show differential expression under oxidative stress. A total of 32 down-regulated putative target genes were also identified among 418 differentially expressed hepatic genes previously reported for the same animal model. Among these, GPR37 (G protein-coupled receptor 37), HEYL (hairy/enhancer-of-split related with YRPW motif-like), DNJA1, CD14 (Cluster of differentiation 14) and GNS (glucosamine (N-acetyl)-6-sulfatase) are known to be associated with hepatic metabolic disorders. In addition miR-140 and miR-2885 have binding sites on the most down-regulated of these genes, FADS2 (Fatty acid desaturase 2) which encodes an enzyme critical in lipid biosynthesis. Furthermore, HNF3-gamma (Hepatocyte nuclear factor 3-gamma), a hepatic transcription factor (TF) that is involved in IGF-1 expression regulation and maintenance of glucose homeostasis is a putative target of miR-31. CONCLUSIONS: This study shows that SNEB affects liver miRNA expression and these miRNAs have putative targets in hepatic genes down-regulated under this condition. This study highlights the potential role of miRNAs in transcription regulation of hepatic gene expression during SNEB in dairy cattle.

The miRNAome of the postpartum dairy cow liver in negative energy balance.

BACKGROUND: Negative energy balance (NEB) is an altered metabolic state in high yielding cows that occurs during the first few weeks postpartum when energy demands for lactation and maintenance exceed the energy supply from dietary intake. NEB can, in turn, lead to metabolic disorders and to reduced fertility. Alterations in the expression of more than 700 hepatic genes have previously been reported in a study of NEB in postpartum dairy cows. miRNAs (microRNA) are known to mediate many alterations in gene expression post transcriptionally. To study the hepatic miRNA content of postpartum dairy cows, including their overall abundance and differential expression, in mild NEB (MNEB) and severe NEB (SNEB), short read RNA sequencing was carried out. To identify putative targets of differentially expressed miRNAs among differentially expressed hepatic genes reported previously in dairy cows in SNEB computational target identification was employed. RESULTS: Our results indicate that the dairy cow liver expresses 53 miRNAs at a lower threshold of 10 reads per million. Of these, 10 miRNAs accounted for greater than 95% of the miRNAome (miRNA content). Of the highly expressed miRNAs, miR-122 constitutes 75% followed by miR-192 and miR-3596. Five out of thirteen let-7 miRNA family members are also among the highly expressed miRNAs. miR-143, down-regulated in SNEB, was found to have 4 putative up-regulated gene targets associated with SNEB including LRP2 (low density lipoprotein receptor-related protein 2), involved in lipid metabolism and up-regulated in SNEB. CONCLUSIONS: This is the first liver miRNA-seq profiling study of moderate yielding dairy cows in the early postpartum period. Tissue specific miR-122 and liver enriched miR-192 are two of the most abundant miRNAs in the postpartum dairy cow liver. miR-143 is significantly down-regulated in SNEB and putative targets of miRNA-143 which are up-regulated in SNEB, include a gene involved in lipid metabolism.