Biology subjects

Georges, M.

Publications and source records attributed to Georges, M..

17 recordsLinked to original sources

bioRxiv2026

Uncertainty-aware benchmarking reveals ambiguous transcripts in mRNA-lncRNA classification

BackgroundLong non-coding RNAs (lncRNAs) have gained significant attention in recent years, yet distinguishing them from protein-coding transcripts remains challenging. Indeed, many lncRNAs share mRNA-like processing and existing sequence-derived signals do not fully capture the coding/non-coding boundary. Recent GENCODE annotation efforts revealed tens of thousands of novel lncRNA sequences as well as the reclassification of some lncRNAs into the protein-coding class, highlighting the need to better characterize transcript features associated with classification uncertainty and errors. ResultsWe performed uncertainty-aware benchmarking by retraining and evaluating eight transcript classifiers under a controlled protocol on a label-stable GENCODE v46-v47 subset. Beyond conventional model evaluation metrics, we quantified inter-tool agreement and entropy-based uncertainty to stratify transcripts into consensus, discordant, and consensus-error groups. To expand standard sequence and ORF-derived signals, we incorporated repeat-derived features from mature transcripts and non-B DNA motif features across gene bodies. Although aggregate performance was high, [~]45% of transcripts showed inter-tool discordance, particularly among lncRNAs. Feature analyses linked low-uncertainty predictions to strong coding-like signals, whereas high-uncertainty profiles exhibited mixed signatures. Alongside classical predictors in global importance analyses, repeat-derived features appear as main contributors. ConclusionsBy combining controlled benchmarking with transcript-level agreement and uncertainty stratification, together with extended feature profiling, we identified patterns associated with classifier disagreement and misclassification. This novel framework provides practical guidance for interpreting predictions, motivating the development of more robust coding/non-coding classifiers, while also shedding light on the sequence properties that distinguish lncRNA sequences.

bioinformatics↗

bioRxiv2026

Reconstructing N-Glycan Profiles of Individual Glycoproteins from the Total Blood Plasma N-Glycome: Implications for Immunoglobulin G N-Glycosylation GWAS

The genetic regulation of the plasma N-glycome variation in human populations is not fully characterized, partly due to the limited sample size in glyco-genetics studies. Here, we aimed to demonstrate that protein-specific N-glycan profiles, like those of immunoglobulin G (IgG), can be accurately reconstructed from the total plasma N-glycome (TPNG), enabling us to find new regulators of this complex process re-analysing existing datasets. By testing multiple linear and non-linear machine learning approaches we built a model to reconstruct IgG N-glycans from TPNG data, training on the TwinsUK cohort and validating on CEDAR. We reconstruct GWAS summary statistics for IgG N-glycans by applying the trained linear model to plasma glycan GWAS summary statistics, i.e., as GWAS of linear combinations of plasma glycan traits. The majority of the identified loci had been implicated in IgG N-glycosylation GWAS. Additionally, we found four new loci and suggested the role of FCRLA, KDELR2, HHEX, and TCF3 in the regulation of IgG N-glycosylation. In conclusion, we showed that our method enables the creation of protein-specific N-glycome datasets, allowing for powerful meta-analyses without the need to profile new samples.

genetics↗

bioRxiv2023

Sequenced-based GWAS for linear classification traits in Belgian Blue beef cattle reveals new coding variants in genes regulating body size in mammals

Cohorts of individuals that have been genotyped and phenotyped for genomic selection programs offer the opportunity to better understand genetic variation associated with complex traits. Here, we perform an association study for traits related to body size and muscular development in intensively selected beef cattle. We leveraged multiple trait information to refine and interpret the significant associations. After a multiple-step genotype imputation to the sequence-level for 14,762 Belgian Blue beef (BBB) cattle cows, we performed a GWAS for 11 traits related to muscular development and body size. The 37 identified genome-wide significant QTL could be condensed in 11 unique QTL regions based on their position. There was evidence for pleiotropic effects in most of these regions (e.g., correlated association signals, overlap between credible sets of candidate variants - CSCV). We consequently applied a multiple-trait approach to combine information from different traits to refine the CSCV. In several QTL regions, we identified strong candidate genes known to be related to growth and height in other species such as LCORL-NCAPG or CCND2. For some of these genes, relevant candidate variants were identified in the CSCV, including three new missense variants in EZH2, PAPPA2 and ADAM12, possibly two additional coding variants in LCORL, and candidate regulatory variants linked to CCND2 and ARMC12. Strikingly, four other QTL regions were related to five (recessive) deleterious coding variants previously identified. Heterozygotes for several of these mutations have favorable effects for muscular development traits. Our study further supports that a set of common genes controls body size across mammalian species. In particular, we added new genes to the list of those associated with height in both human and cattle. We also identified new strong candidate causing variants in some of those genes, strengthening the evidence of the causality of these genes. Several breed-specific recessive deleterious variants were identified in our QTL regions, probably as a result of the extreme selection for muscular development in BBB cattle.

genetics↗

bioRxiv2023

Nanopore long-reads reveal fine structure of prokaryotic communities in mangrove sediments, like Illumina short-reads but with twice more taxa

Following the development of high-throughput DNA sequencers, environmental prokaryotic communities were usually described by metabarcoding on short markers of the 16S domain. Among third generation sequencers, that offered the possibility to sequence the full 16s domain, the portable MinION from Oxford Nanopore was undervalued for metabarcoding because of its relatively higher error rate per read. Here we illustrate the limits and benefits of Nanopore sequencing devices by comparing the prokaryotic community structure in a mock community and 52 sediment samples from mangrove sites, inferred from full-length 16S long-reads (16S-FL, ca. 1.5 kpb) on a MinION device, with those inferred from partial 16S short-reads (16S- V4V5, ca. 0.4kpb, 16S-V4V5) on Illumina MiSeq. 16S-V4V5 and 16S-FL retrieved all the bacterial species from the mock, but Nanopore long-reads overestimated their diversity more than twice. Whether these supplementary OTUs were artefactual or not, they only accounted for ca. 10% of the reads. From the sediment samples, with a coverage-based rarefaction of reads and after singletons filtering, Mantel and Procrustean tests of co-inertia showed that bacterial community structures inferred from 16S-V4V5 and 16S- FL were significantly similar, showing both a comparable contrast between sites and a coherent sea-land orientation within sites. In our dataset, 84.7 and 98.8% of the 16S-V4V5 assigned reads were assigned strictly to the same species and genus, respectively, than those detected by 16S-FL. 16S-FL allowed to detect 92.2% of the 309 families and 87.7% of the 448 genera that were detected by the short 16S-V4V5. 16S-FL recorded 973 additional species and 392 genus not detected by 16S-V4V5 (31.5 and 10.4% of the 16S-FL reads, respectively, among which 67.8 and 79.3% were assigned), producted by both primer specificities and diffrent error rates. Thus, our results concluded to an overall similarity between 16S-V4V5 and 16S-FL sequencing strategies for this type of environmental samples.

molecular biology↗

bioRxiv2022

Lack of evidence supporting transgenerational effects of non-transmitted paternal alleles on the murine transcriptome

Transgenerational genetic effects are defined as the effects of untransmitted parental alleles on the phenotype of their offspring. Well-known transgenerational genetic effects, in humans and other mammals, are the effects of a parental genotype on the nurturing ability of the parents, coined "genetic nurture". However, there exist examples of transgenerational genetic effects in model organisms that are independent of nurturing effects and support the epigenetic transmission of a memory of the parental genotype possibly mediated by small RNA species. To test whether such transgenerational epigenetic effects might exist in mammals, we generated 833 isogenic C57BL/6J (B6) mice that differed only by the presence in the genome of their sire of one copy of four A/J chromosomes (MMU 15, 17, 19 or X). We measured 25 anatomical traits and performed RNA-Seq on five distinct tissues (heart, liver, pituitary, whole embryo, and placenta). There was no evidence of a significant effect from untransmitted A/J sire chromosome alleles, whether on anatomical traits or gene expression level. We observed an effect on Mid1 expression levels in multiple tissues, but this was shown to be due to a de novo mutation that occurred in one of the sire lines. We conclude that transgenerational epigenetic memory of non-transmitted paternal alleles - if it exists - is uncommon in mice and likely other mammals.

genetics↗

bioRxiv2022

Dual blockade of misfolded alpha-sarcoglycan degradation by bortezomib and givinostat combination

Limb-girdle muscular dystrophy type R3 (LGMD R3) is a rare genetic disorder characterized by a progressive proximal muscle weakness and caused by mutations in the SGCA gene encoding alpha-sarcoglycan (-SG). Here, we report the results of a mechanistic screening ascertaining the molecular mechanisms involved in the degradation of the most prevalent misfolded R77C--SG protein. We performed a combinatorial study to identify drugs potentializing the effect of a low dose of the proteasome inhibitor bortezomib on the R77C--SG degradation inhibition. Analysis of the screening associated to artificial intelligence-based predictive ADMET characterization of the hits led to identification of the HDAC inhibitor givinostat as potential therapeutical candidate. Functional characterization revealed that givinostat effect was related to autophagic pathway inhibition, unveiling new theories concerning degradation pathways of misfolded SG proteins. Beyond the identification of a new therapeutic option for LGMD R3 patients, our results shed light on the potential repurposing of givinostat for the treatment of other genetic diseases sharing similar protein degradation defects such as LGMD R5 and cystic fibrosis.

pharmacology and toxicology↗

bioRxiv2021

Benchmarking phasing software with a whole-genome sequenced cattle pedigree

BackgroundAccurate haplotype reconstruction is required in many applications in quantitative and population genomics. Different phasing methods are available but their accuracy must be evaluated for samples with different properties (population structure, marker density, etc.). We herein took advantage of whole-genome sequence data available for a Holstein cattle pedigree containing 264 individuals, including 98 trios, to evaluate several population-based phasing methods. This data represents a typical example of a livestock population, with low effective population size, high levels of relatedness and long-range linkage disequilibrium. ResultsAfter stringent filtering of our sequence data, we evaluated several population-based phasing programs including one or more versions of AlphaPhase, ShapeIT, Beagle, Eagle and FImpute. To that end we used 98 individuals having both parents sequenced for validation. Their haplotypes reconstructed based on Mendelian segregation rules were considered the gold standard to assess the performance of population-based methods in two scenarios. In the first one, only these 98 individuals were phased, while in the second one, all the 264 sequenced individuals were phased simultaneously, ignoring the pedigree relationships. We assessed phasing accuracy based on switch error counts (SEC) and rates (SER), lengths of correctly phased haplotypes and pairwise SNP phasing accuracies (the probability that a pair of SNPs is correctly phased as a function of their distance). For most evaluated metrics or scenarios, the best software was either ShapeIT4.1 or Beagle5.2, both methods resulting in particularly high phasing accuracies. For instance, ShapeIT4.1 achieved a median SEC of 50 per individual and a mean haplotype block length of 24.1 Mb in the second scenario. These statistics are remarkable since the methods were evaluated with a map of 8,400,000 SNPs, and this corresponds to only one switch error every 40,000 phased informative markers. When more relatives were included in the data, FImpute3.0 reconstructed extremely long segments without errors. ConclusionsWe report extremely high phasing accuracies in a typical livestock sample of 100 sequenced individuals. ShapeIT4.1 and Beagle5.2 proved to be the most accurate, particularly for phasing long segments. Nevertheless, most tools achieved high accuracy at short distances and would be suitable for applications requiring only local haplotypes.

genomics↗

bioRxiv2021

The swine spatiotemporal H3K27ac spectrum provides novel resources for exploring gene regulation related to complex traits and fundamental biological process

The limited knowledge of genomic non-coding and regulatory regions has limited our ability to decipher the genetic mechanisms underlying complex traits in pigs. In this study, we characterize the spatiotemporal landscape of putative enhancers and promoters and their target genes by combining H3K27ac targeted ChIP-Seq and RNA-Seq in fetal (day 74-75 pc) and adult (day 132-150 pn) tissues (brain, liver, heart, muscle and small intestine) sampled from Asian aboriginal Bamaxiang and European highly selected Large White pigs of both sexes. We identify 101,290 H3K27ac peaks marking 18,521 promoters and 82,769 enhancers, including peaks that are active across all tissues and developmental stages could indicate safe harbors for exogenous gene insertion, and tissue and developmental-stage specific peaks that regulate genes pathways matching tissue and developmental stage specific physiological functions. We found H3K27ac and DNA methylation in the promoter region of the XIST gene may involve in X chromosome inactivation, and demonstrate utility of the present resource to reveal regulatory patterns of known causal genes and to prioritize candidate causal variants for complex traits in pigs. We have developed a web browser to improve the accessibility of the results (http://39.108.231.116/browser/?genome=susScr11).

genomics↗

bioRxiv2021

Identification of tissue-specific and common methylation quantitative trait loci in healthy individuals using MAGAR

BackgroundUnderstanding the influence of genetic variants on DNA methylation is fundamental for the interpretation of epigenomic data in the context of disease. There is a need for systematic approaches not only for determining methylation quantitative trait loci (methQTL) but also for discriminating general from cell-type-specific effects. ResultsHere, we present a two-step computational framework MAGAR, which fully supports identification of methQTLs from matched genotyping and DNA methylation data, and additionally the identification of quantitative cell-type-specific methQTL effects. In a pilot analysis, we apply MAGAR on data in four tissues (ileum, rectum, T-cells, B-cells) from healthy individuals and demonstrate the discrimination of common from cell-type-specific methQTLs. We experimentally validate both types of methQTLs in an independent dataset comprising additional cell types and tissues. Finally, we validate selected methQTLs (PON1, ZNF155, NRG2) by ultra-deep local sequencing. In line with previous reports, we find cell-type-specific methQTLs to be preferentially located in enhancer elements. ConclusionsOur analysis demonstrates that a systematic analysis of methQTLs provides important new insights on the influences of genetic variants to cell-type-specific epigenomic variation.

genetics↗

bioRxiv2021

A 12 kb multi-allelic copy number variation encompassing a GC gene enhancer is associated with mastitis resistance in dairy cattle

Clinical mastitis (CM) is an inflammatory disease occurring in the mammary glands of lactating cows. CM is under genetic control, and a prominent CM resistance QTL located on chromosome 6 was reported in various dairy cattle breeds. Nevertheless, the biological mechanism underpinning this QTL has been lacking. Herein, we mapped, fine-mapped, and discovered the putative causal variant underlying this CM resistance QTL in the Dutch dairy cattle population. We identified a ~12 kb multi-allelic copy number variant (CNV), that is in perfect linkage disequilibrium with a GWAS lead SNP, as a promising candidate variant. By implementing a genome-wide association study (GWAS) and through expression QTL mapping, we showed that the group-specific component gene (GC), a gene encoding a vitamin D binding protein, is an excellent candidate causal gene for the QTL. The multiplicated alleles are associated with increased GC expression and low CM resistance. Ample evidence from functional genomics data supports the presence of an enhancer within this CNV, which would exert cis-regulatory effect on GC. We observed that strong positive selection swept the region near the CNV, and haplotypes associated with the multiplicated allele were strongly selected for. Moreover, the multiplicated allele showed pleiotropic effects for increased milk yield and reduced fertility, hinting that a shared underlying biology for these effects may revolve around the vitamin D pathway. These findings together suggest a putative causal variant of a CM resistance QTL, where a cis-regulatory element located within a CNV can alter gene expression and affect multiple economically important traits. Author summaryClinical mastitis (CM) is an inflammatory disease that negatively influences dairy production and compromises animal welfare. Although one major genetic locus for CM resistance was mapped on bovine chromosome 6, a mechanistic description of this association has been lacking. Herein, we report a 12-kb multiallelic copy number variant (CNV), encompassing a strong enhancer for group-specific component gene (GC), as a likely causal variant for this locus. This CNV is associated with high GC expression and low CM resistance. We speculate that upregulation of GC leads to a large amount of vitamin D binding protein, which in turn, reduces biologically available vitamin D, resulting in vitamin D deficiency and low CM resistance. Despite the negative effect on CM resistance, the CNV contributes to increased milk production, hinting at balancing selection. Our results highlight how multiplication of a regulatory element can shape economically important traits in dairy cattle, both in favourable and unfavourable directions.

genomics↗

bioRxiv2020

An ancient deletion in the ABO gene affects the composition of the porcine microbiome by altering intestinal N-acetyl-galactosamine concentrations

We have generated a large heterogenous stock population by intercrossing eight divergent pig breeds for multiple generations. We have analyzed the composition of the intestinal microbiota at different ages and anatomical locations in > 1,000 6th- and 7th- generation animals. We show that, under conditions of exacerbated genetic yet controlled environmental variability, microbiota composition and abundance of specific taxa (including Christensenellaceae) are heritable in this monogastric omnivore. We fine-map a QTL with major effect on the abundance of Erysipelotrichaceae to chromosome 1q and show that it is caused by a common 2.3-Kb deletion inactivating the ABO acetyl-galactosaminyl-transferase gene. We show that this deletion is a trans-species polymorphism that is [≥]3.5 million years old and under balancing selection. We demonstrate that it acts by decreasing the concentrations of N-acetyl-galactosamine in the cecum thereby reducing the abundance of Erysipelotrichaceae strains that have the capacity to import and catabolize N-acetyl-galactosamine.

genetics↗

bioRxiv2020

Combined analysis of single cell RNA-Seq and ATAC-Seq data reveals regulatory toggles operating in native and iPS-derived retina.

We report the generation and analysis of single-cell RNA-Seq data (> 38,000 cells) from native and iPSC-derived murine retina at four matched developmental stages spanning the emergence of the major retinal cell types. We combine information from temporal sampling, visualization of 3D UMAP manifolds, pseudo-time and RNA velocity analyses, to show that iPSC-derived 3D retinal aggregates broadly recapitulate the native developmental trajectories. However, we show relaxation of spatial and temporal transcriptome control, premature emergence and dominance of photoreceptor precursor cells, and susceptibility of dynamically regulated pathways and transcription factors to culture conditions in iPSC-derived retina. We generate bulk ATAC-Seq data for native and iPSC-derived murine retina identifying [~]125,000 peaks. We combine single-cell RNA-Seq with ATAC-Seq information and obtain evidence that approximately half the transcription factors that are dynamically regulated during retinal development may act as repressors rather than activators. We propose that sets of activators and repressors with cell-type specific expression constitute "regulatory toggles" that lock cells in distinct transcriptome states underlying differentiation. We provide evidence supporting our hypothesis from the analysis of publicly available single-cell ATAC-Seq data for adult mouse retina. We identify subtle but noteworthy differences in the operation of such toggles between native and iPSC-derived retina particularly for the Etv1, Etv5, Hes1 and Zbtb7a group of transcription factors.

developmental biology↗

bioRxiv2020

Genetic Analyses of Blood Cell Structure for Biological and Pharmacological Inference

Thousands of genetic associations with phenotypes of blood cells are known, but few are with phenotypes relevant to cell function. We performed GWAS of 63 flow-cytometry phenotypes, including measures of cell granularity, nucleic acid content, and reactivity, in 39,656 participants in the INTERVAL study, identifying 2,172 variant-trait associations. These include associations mediated by functional cellular structures such as secretory granules, implicated in vascular, thrombotic, inflammatory and neoplastic diseases. By integrating our results with epigenetic data and with signals from molecular abundance/disease GWAS, we infer the hematopoietic origins of population phenotypic variation and identify the transcription factor FOG2 as a regulator of platelet -granularity. We show how flow cytometry genetics can suggest cell types mediating complex disease risk and suggest efficacious drug targets, presenting Daclizumab/Vedolizumab in autoimmune disease as positive controls. Finally, we add to existing evidence supporting IL7/IL7-R as drug targets for multiple sclerosis.

genetics↗

bioRxiv2019

SNP-based quantitative deconvolution of biological mixtures: application to the detection of cows with subclinical mastitis by whole genome sequencing of tank milk.

Biological products of importance in food (f.i. milk) and medical (f.i. donor blood derived products) sciences often correspond to mixtures of samples contributed by multiple individuals. Identifying which individuals contributed to the mixture and in what proportions may be of interest in several circumstances. We herein present a method that allows to do this by shallow whole genome sequencing of the DNA in mixed samples from hundreds of donors. We demonstrate the efficacy of the approach for the detection of cows with subclinical mastitis by analysis of farms tank mixtures containing milk from as many as 500 cows.

genomics↗

bioRxiv2019

Pooled CRISPR Inverse PCR sequencing (PCIP-seq): simultaneous sequencing of retroviral insertion points and the associated provirus in thousands of cells with long reads

Retroviral infections create a large population of cells, each defined by a unique proviral insertion site. Methods based on short-read high throughput sequencing can identify thousands of insertion sites, but the proviruses within remain unobserved. We have developed Pooled CRISPR Inverse PCR sequencing (PCIP-seq), a method that leverages long reads on the Oxford Nanopore MinION platform to sequence the insertion site and its associated provirus. We have applied the technique to three exogenous retroviruses, HTLV-1, HIV-1 and BLV, as well as endogenous retroviruses in both cattle and sheep. The long reads of PCIP-seq improved the accuracy of insertion site identification in repetitive regions of the genome. The high efficiency of the method facilitated the identification of tens of thousands of insertion sites in a single sample. We observed thousands of SNPs and dozens of structural variants within proviruses and uncovered evidence of viral hypermutation, recombination and recurrent selection.

genomics↗

bioRxiv2018

Unraveling the polygenic architecture of complex traits using blood eQTL meta-analysis

SummaryWhile many disease-associated variants have been identified through genome-wide association studies, their downstream molecular consequences remain unclear.\n\nTo identify these effects, we performed cis- and trans-expression quantitative trait locus (eQTL) analysis in blood from 31,684 individuals through the eQTLGen Consortium.\n\nWe observed that cis-eQTLs can be detected for 88% of the studied genes, but that they have a different genetic architecture compared to disease-associated variants, limiting our ability to use cis-eQTLs to pinpoint causal genes within susceptibility loci.\n\nIn contrast, trans-eQTLs (detected for 37% of 10,317 studied trait-associated variants) were more informative. Multiple unlinked variants, associated to the same complex trait, often converged on trans-genes that are known to play central roles in disease etiology.\n\nWe observed the same when ascertaining the effect of polygenic scores calculated for 1,263 genome-wide association study (GWAS) traits. Expression levels of 13% of the studied genes correlated with polygenic scores, and many resulting genes are known to drive these traits.

genomics↗

bioRxiv2018

Defining the genetic control of human blood plasma N-glycome using genome-wide association study

Glycosylation is a common post-translational modification of proteins. It is known, that glycans are directly involved in the pathophysiology of every major disease. Defining genetic factors altering glycosylation may provide a basis for novel approaches to diagnostic and pharmaceutical applications. Here, we report a genome-wide association study of the human blood plasma N-glycome composition in up to 3811 people. We discovered and replicated twelve loci. This allowed us to demonstrate a clear overlap in genetic control between total plasma and IgG glycosylation. Majority of loci contained genes that encode enzymes directly involved in glycosylation (FUT3/FUT6, FUT8, B3GAT1, ST6GAL1, B4GALT1, ST3GAL4, MGAT3, and MGAT5). We, however, also found loci that are likely to reflect other, more complex, aspects of plasma glycosylation process. Functional genomic annotation suggested the role of DERL3, which potentially highlights the role of glycoprotein degradation pathway, and such transcription factor as IKZF1.

genetics↗