Biology subjects

Guelfi, S.

Publications and source records attributed to Guelfi, S..

13 recordsLinked to original sources

bioRxiv2023

Systematic visualisation of molecular QTLs reveals variant mechanisms at GWAS loci

Splicing quantitative trait loci (QTLs) have been implicated as a common mechanism underlying complex trait associations. However, utilising splicing QTLs in target discovery and prioritisation has been challenging due to extensive data normalisation which often renders the direction of the genetic effect as well as its magnitude difficult to interpret. This is further complicated by the fact that strong expression QTLs often manifest as weak splicing QTLs and vice versa, making it difficult to uniquely identify the underlying molecular mechanism at each locus. We find that these ambiguities can be mitigated by visualising the association between the genotype and average RNA sequencing read coverage in the region. Here, we generate these QTL coverage plots for 1.7 million molecular QTL associations in the eQTL Catalogue identified with five quantification methods. We illustrate the utility of these QTL coverage plots by performing colocalisation between vitamin D levels in the UK Biobank and all molecular QTLs in the eQTL Catalogue. We find that while visually confirmed splicing QTLs explain just 6/53 of the colocalising signals, they are significantly less pleiotropic than eQTLs and identify a prioritised causal gene in 4/6 cases. All our association summary statistics and QTL coverage plots are freely available at https://www.ebi.ac.uk/eqtl/.

genomics↗

bioRxiv2023

Splicing accuracy varies across human introns, tissues and age

Alternative splicing impacts most multi-exonic human genes. Inaccuracies during this process may have an important role in ageing and disease. Here, we investigated mis-splicing using RNA-sequencing data from ~14K control samples and 42 human body sites, focusing on split reads partially mapping to known transcripts in annotation. We show that mis-splicing occurs at different rates across introns and tissues and that these splicing inaccuracies are primarily affected by the abundance of core components of the spliceosome assembly and its regulators. Using publicly available data on short-hairpin RNA-knockdowns of numerous spliceosomal components and related regulators, we found support for the importance of RNA-binding proteins in mis-splicing. We also demonstrated that age is positively correlated with mis-splicing, and it affects genes implicated in neurodegenerative diseases. This in-depth characterisation of mis-splicing can have important implications for our understanding of the role of splicing inaccuracies in human disease and the interpretation of long-read RNA-sequencing data.

genetics↗

bioRxiv2023

Unravelling cell type specific response to Parkinson's Disease at single cell resolution

Parkinsons Disease (PD) is the second most common neurodegenerative disorder and is generally characterized by impaired motor functions. It currently affects 6.3 million people aged 60 years and more, worldwide. The pathological hallmarks of PD are Lewy bodies (abnormal aggregation of -synuclein inside cells), which are observed primarily in the substantia nigra (SN) region of the midbrain. It is yet not known how different cell types in SN respond during PD and what are the molecular mechanisms underlying neurodegeneration. To address this question, we generated a large-scale single cell transcriptomics dataset from human post-mortem SN tissue of 29 donors including 15 sporadic cases and 14 controls. We obtained data for a total of [~]80K nuclei, representing major cell types of the brain (including neurons, astrocytes, microglia and oligodendrocytes). Pathway and differential gene expression analysis revealed multicellular character of PD pathology involving major cellular response from neuronal and glial cells.

bioinformatics↗

bioRxiv2022

Single-nucleus co-expression networks of dopaminergic neurons support iron accumulation as a plausible explanation to their vulnerability in Parkinson's disease

Motivationgene co-expression networks have been widely applied to identify critical genes and pathways for neurodegenerative diseases such as Parkinsons and Alzheimers disease. Now, with the advent of single-cell RNA-sequencing, we have the opportunity to create cell-type specific gene co-expression networks. However, single-cell RNA-sequencing data is characterized by its sparsity, amongst some other issues raised by this new type of data. ResultsWe present scCoExpNets, a framework for the discovery and analysis of cell-type specific gene coexpression networks (GCNs) from single-cell RNA-seq data. We propose a new strategy to address the problem of sparsity, named iterative pseudo-cell identification. It consists of adding the gene expression of pairs of cells that belong to the same individual and the same cell-type while the number of cells is over 200, thus creating multiple matrices and multiple scGCNs for the same cell-type, all of them seen as alternative and complementary views of the same phenomena. We applied this new tool on a snRNA-seq dataset human post-mortem substantia nigra pars compacta tissue of 13 controls and 14 Parkinsons disease (PD) cases (18 males and 9 females) with 30-99 years. We show that one of the hypotheses that support the selective vulnerability of dopaminergic neurons in PD, the iron accumulation, is sustained in our dopaminergic neurons network models. Moreover, after successive pseudo-celluling iterations, the gene groups sustaining this hypothesis remain intact. At the same time, this pseudo-celulling strategy also allows us to discover genes whose grouping changes considerably throughout the iterations and provides new insights. Finally, since some of our models were correlated with diagnosis and age at the same time, we also developed our own framework to create covariate-specific GCNs, called CovCoExpNets. We applied this new software to our snRNA-seq dataset and we identified 11 age-specific genes and 5 diagnosis-specific genes which do not overlap. Availability and implementationThe CoExpNets implementations are available as R packages: scCoExpNets for creating single-cell GCNs and CovCoExpNets for creating covariate-specific GCNs. Users can either download the development version via github https://github.com/aliciagp/scCoExpNets and https://github.com/aliciagp/CovCoExpNets Contactalicia.gomez1@um.es Supplementary informationsupplementary data is available online.

bioinformatics↗

bioRxiv2022

Analysis of nucleus and cytoplasm-specific RNA fractions demonstrates that a significant proportion of the genetic regulation of gene expression across the human brain occurs post-transcriptionally

Gaining insight into the genetic regulation of gene expression in human brain is key to the interpretation of genome-wide association studies for major neurological and neuropsychiatric diseases. Expression quantitative trait loci (eQTL) analyses have largely been used to achieve this, providing valuable insights into the genetic regulation of steady-state RNA in human brain, but not distinguishing between molecular processes regulating transcription and stability. RNA quantification within cellular fractions can disentangle these processes in cell types and tissues which are challenging to model in vitro. We investigated the underlying molecular processes driving the genetic regulation of gene expression specific to a cellular fraction using allele-specific expression (ASE). Applying ASE analysis to genomic and transcriptomic data from paired nuclear and cytoplasmic fractions of anterior prefrontal cortex, cerebellar cortex and putamen tissues from 4 post-mortem neuropathologically-confirmed control human brains, we demonstrate that a significant proportion of genetic regulation of gene expression occurs post-transcriptionally in the cytoplasm, with genes undergoing this form of regulation more likely to be synaptic. These findings have implications for understanding the structure of gene expression regulation in human brain, and importantly the interpretation of rapidly growing single-nucleus brain RNA-sequencing and eQTL datasets, where cytoplasm-specific regulatory events could be missed.

genomics↗

bioRxiv2021

Leveraging omic features with F3UTER enables identification of unannotated 3'UTRs for synaptic genes

There is growing evidence for the importance of 3 untranslated region (3UTR) dependent regulatory processes. However, our current human 3UTR catalogue is incomplete. Here, we developed a machine learning-based framework, leveraging both genomic and tissue-specific transcriptomic features to predict previously unannotated 3UTRs. We identify unannotated 3UTRs associated with 1,513 genes across 39 human tissues, with the greatest abundance found in brain. These unannotated 3UTRs were significantly enriched for RNA binding protein (RBP) motifs and exhibited high human lineage-specificity. We found that brain-specific unannotated 3UTRs were enriched for the binding motifs of important neuronal RBPs such as TARDBP and RBFOX1, and their associated genes were involved in synaptic function and brain- related disorders. Our data is shared through an online resource F3UTER (https://astx.shinyapps.io/F3UTER/). Overall, our data improves 3UTR annotation and provides novel insights into the mRNA-RBP interactome in the human brain, with implications for our understanding of neurological and neurodevelopmental diseases.

bioinformatics↗

bioRxiv2020

Regulation of mitophagy by the NSL complex underlies genetic risk for Parkinson's disease at Chr16q11.2 and on the MAPT H1 allele.

Parkinsons disease (PD) is a common incurable neurodegenerative disease. The identification of genetic variants via genome-wide association studies (GWAS) has considerably advanced our understanding of the PD genetic risk. Understanding the functional significance of the risk loci is now a critical step towards translating these genetic advances into an enhanced biological understanding of the disease. Impaired mitophagy is a key causative pathway in familial PD, but its relevance to idiopathic PD is unclear. We used a mitophagy screening assay to evaluate the functional significance of risk genes identified through GWAS. We identified two new regulators of PINK1-mitophagy, KAT8 and KANSL1, previously shown to modulate lysine acetylation. We show that KAT8 and KANSL1 modulate PINK1 gene expression and subsequent PINK1-mitophagy. These findings suggest PINK1-mitophagy is a contributing factor to idiopathic PD. KANSL1 is located on chromosome 17q21 where the risk associated gene has long been considered to be MAPT. While our data does not exclude a possible association between the MAPT gene and PD, it provides strong evidence that KANSL1 plays a crucial role in the disease. Finally, these results enrich our understanding of physiological events regulating mitophagy and establish a novel pathway for drug targeting in neurodegeneration.

neuroscience↗

bioRxiv2019

Integration of eQTL and Parkinson’s disease GWAS data implicates 11 disease genes

Substantial genome-wide association study (GWAS) work in Parkinsons disease (PD) has led to an increasing number of loci shown reliably and robustly to be associated with the increased risk of the disease. Prioritising causative genes and pathways from these studies has proven problematic. Here, we present a comprehensive analysis of PD GWAS data with expression and methylation quantitative trait loci (eQTL/mQTL) using Colocalisation analysis (Coloc) and transcriptome-wide association analysis (TWAS) to uncover putative gene expression and splicing mechanisms driving PD GWAS signals. Candidate genes were further characterised by determining cell-type specificity, weighted gene co-expression (WGNCA) and protein-protein interaction (PPI) networks.\n\nGene-level analysis of expression revealed 5 genes (WDR6, CD38, GPNMB, RAB29, TMEM163) that replicated using both Coloc and TWAS analyses in both GTEx and Braineac expression datasets. A further 6 genes (ZRANB3, PCGF3, NEK1, NUPL2, GALC, CTSB) showed evidence of disease-associated splicing effects. Cell-type specificity analysis revealed that gene expression was overall more prevalent in glial cell-types compared to neurons. The WGNCA analysis showed that NUPL2 is a key gene in 3 modules implicated in catabolic processes related with protein ubiquitination (protein ubiquitination (p=7.47e-10) and ubiquitin-dependent protein catabolic process (p = 2.57e-17) in nucleus accumbens, caudate and putamen, while TMEM163 and ZRANB3 were both important in modules indicating regulation of signalling (p=1.33e-65] and cell communication (p=7.55e-35) in the frontal cortex and caudate respectively. PPI analysis and simulations using random networks demonstrated that the candidate genes interact significantly more with known Mendelian PD and parkinsonism proteins than would be expected by chance. The proteins core proteins this network were enriched for regulation of the ERBB receptor tyrosine protein kinase signalling pathways.\n\nTogether, these results point to a number of candidate genes and pathways that are driving the associations observed in PD GWAS studies.

neuroscience↗

bioRxiv2019

ERASE: Extended Randomization for assessment of annotation enrichment in ASE datasets

Genome-wide association studies (GWAS) have identified thousands of genetic variants associated with various human phenotypes and many of these loci are thought to act at a molecular level by regulating gene expression. Detection of allele specific expression (ASE), namely preferential usage of an allele at a transcribed locus, is an increasingly important means of studying the genetic regulation of gene expression. However, there are currently a paucity of tools available to link ASE sites with GWAS risk loci. Existing integration methods first use ASE sites to infer cis-acting expression quantitative trait loci (eQTL) and then apply eQTL-based approaches. ERASE is a method that assesses the enrichment of risk loci amongst ASE sites directly. Furthermore, ERASE enables additional biological insights to be made through the addition of other SNP level annotations. ERASE is based on a randomization approach and controls for read depth, a significant confounder in ASE analyses. In this paper, we demonstrate that ERASE can efficiently detect the enrichment of eQTLs and risk loci within ASE data and that it remains sensitive even when used with underpowered GWAS datasets. Finally, using ERASE in combination with GWAS data for Parkinsons disease and data on the splicing potential of individual SNPs, we provide evidence to suggest that risk loci for Parkinsons disease are enriched amongst ASEs likely to affect splicing. Thus, we show that ERASE is an important new tool for the integration of ASE and GWAS data, capable of providing novel insights into the pathophysiology of complex diseases.

bioinformatics↗

bioRxiv2019

Regulatory sites for known and novel splicing in human basal ganglia are enriched for disease-relevant information

Genome-wide association studies have generated an increasing number of common genetic variants that affect neurological and psychiatric disease risk. Given that many causal variants are likely to operate by regulating gene expression, an improved understanding of the genetic control of gene expression in human brain is vital. However, the difficulties of sampling human brain, and its complexity, has meant that brain-related expression quantitative trait loci (eQTL) and allele specific expression (ASE) signals have been more limited in their explanatory power than might otherwise be expected. To address this, we use paired genomic and transcriptomic data from putamen and substantia nigra dissected from 117 brains, combined with a comprehensive set of analyses, to interrogate regulation at different stages of RNA processing and uncover novel transcripts. We identify disease-relevant regulatory loci and reveal the types of analyses and regulatory positions yielding the most disease-specific information. We find that splicing eQTLs are enriched for neuron-specific regulatory information; that ASE analyses provide highly cell-specific regulatory information; and that incomplete annotation of the brain transcriptome limits the interpretation of risk loci for neuropsychiatric disease. We release this rich resource of regulatory data through a searchable webserver, http://braineacv2.inf.um.es/.

bioinformatics↗

bioRxiv2018

Incomplete annotation of OMIM genes is likely to be limiting the diagnostic yield of genetic testing, particularly for neurogenetic disorders

Although the increasing use of whole-exome and whole-genome sequencing have improved the yield of genetic testing for Mendelian disorders, an estimated 50% of patients still leave the clinic without a genetic diagnosis. This can be attributed in part to our lack of ability to accurately interpret the genetic variation detected through next-generation sequencing. Variant interpretation is fundamentally reliant on accurate and complete gene annotation, however numerous reports and discrepancies between gene annotation databases reveals that the knowledge of gene annotation remains far from comprehensive. Here, we detect and validate transcription in an annotation-agnostic manner across all 41 different GTEx tissues, then connect novel transcription to known genes, ultimately improving the annotation of 63% of the known OMIM-morbid genes. We find the majority of novel transcription to be tissue-specific in origin, with brain tissues being most susceptible to misannotation. Furthermore, we find that novel transcribed regions tend to be poorly conserved, but are significantly depleted for genetic variation within humans, suggesting they are functionally significant and potentially have human-specific functions. We present our findings through an online platform vizER, which enables individual genes to be visualised and queried for evidence of misannotation. We also release all tissue-specific transcriptomes in a BED format for ease of integration with whole-genome sequencing data. We anticipate that these resources will improve the diagnostic yield for a wide range of Mendelian disorders.

bioinformatics↗

bioRxiv2018

G2P: Using machine learning to understand and predict genes causing rare neurological disorders

To facilitate precision medicine and neuroscience research, we developed a machine-learning technique that scores the likelihood that a gene, when mutated, will cause a neurological phenotype. We analysed 1126 genes relating to 25 subtypes of Mendelian neurological disease defined by Genomics England (March 2017) together with 154 gene-specific features capturing genetic variation, gene structure and tissue-specific expression and co-expression. We randomly re-sampled genes with no known disease association to develop bootstrapped decision-tree models, which were integrated to generate a decision tree-based ensemble for each disease subtype. Genes generating larger numbers of distinct transcripts and with higher probability of having missense mutations in normal individuals were significantly more likely to cause neurological diseases. Using mouse-mutant phenotypic data we tested the accuracy of gene-phenotype predictions and found that for 88% of all disease subtypes there was a significant enrichment of relevant phenotypic abnormalities when predicted genes were mutated in mice and in many cases mutations produced specific and matching phenotypes. Furthermore, using only newly identified genes included in the Genomics England November 2017 release, we assessed our gene-phenotype predictions and showed an 8.3 fold enrichment relative to chance for correct predictions. Thus, we demonstrate both the explanatory and predictive power of machine-learning-based models in neurological disease.

genetics↗

bioRxiv2017

Amygdalar atrophy as the genetically mediated hub of limbic degeneration in Alzheimer’s disease

Pharmacological progress, basic science and medical practice can benefit from objective biomarkers that assist in early diagnosis and prognostic stratification of diseases. In the field of Alzheimers disease (AD), the clinical presentation of early stage dementia may not fulfill any diagnostic criteria for years, and quantifying structural brain changes by magnetic resonance imaging (MRI) has shown promise in the discovery of sensitive biomarkers. Although hippocampal atrophy is often used as an in vivo measure of AD, data-driven neuroimaging has revealed complex patterns of regional brain vulnerability that may not perfectly map to anatomical boundaries. In addition to aiding diagnosis, decoding genetic influences on neuroimaging measures of the disease can enlighten molecular mechanisms of the underlying pathology in living patients and guide the therapeutic design.\n\nHere, we aimed to extract a data-driven MRI feature of brain atrophy in AD by decomposing structural neuroimages using independent component analysis (ICA), a method for performing unbiased computational search in high dimensional data spaces. Our study of the AD Neuroimaging Initiative dataset (n=1,100 subjects) revealed a disease-vulnerable feature with a network-like topology, comprising amygdala, hippocampus, fornix and the inter-connecting white-matter tracts of the limbic system. Whole-genome sequencing identified a nonsynonymous variant (rs34173062) in SHARPIN, a gene coding for a synaptic protein, as a significant modifier of this new MRI feature (p=2.1x10-10). The risk variant was brought to replication in the UK Biobank dataset (n=8,428 subjects), where it was associated with reduced cortical thickness in areas co-localizing with those of the discovery sample (left entorhinal cortex p=0.002, right entorhinal cortex p=8.6x10-4; same direction), as well as with the history of AD in both parents (p=2.3x10-6; same direction).\n\nIn conclusion, our study shows that ICA can transform voxel-wise volumetric measures of the brain into a data-driven feature of neurodegeneration in AD. Structure of the limbic system, as the most vulnerable focus of brain atrophy in AD, is affected by genetic variability of SHARPIN. The elevated risk of dementia in carriers of the minor allele supports engagement of SHARPIN in the disease pathways, and its role in neurotransmitter receptor scaffolding and integrin signaling may inform on new molecular mechanisms of AD pathophysiology.\n\nAbbreviationsAlzheimers disease (AD), genome-wide association study (GWAS), independent component analysis (ICA), mild cognitive impairment (MCI), medial temporal circuit (MTC), single-nucleotide polymorphism (SNP), tensor-based morphometry (TBM)

neuroscience↗