Biology subjects

Fellay, J.

Publications and source records attributed to Fellay, J..

35 records · Page 2Linked to original sources

bioRxiv2021

Truly Privacy-Preserving Federated Analytics for Precision Medicine with Multiparty Homomorphic Encryption

Using real-world evidence in biomedical research, an indispensable complement to clinical trials, requires access to large quantities of patient data that are typically held separately by multiple healthcare institutions. Centralizing those data for a study is often infeasible due to privacy and security concerns. Federated analytics is rapidly emerging as a solution for enabling joint analyses of distributed medical data across a group of institutions, without sharing patient-level data. However, existing approaches either provide only limited protection of patients privacy by requiring the institutions to share intermediate results, which can in turn leak sensitive patient-level information, or they sacrifice the accuracy of results by adding noise to the data to mitigate potential leakage. We propose FAMHE, a novel federated analytics system that, based on multiparty homomorphic encryption (MHE), enables privacy-preserving analyses of distributed datasets by yielding highly accurate results without revealing any intermediate data. We demonstrate the applicability of FAMHE to essential biomedical analysis tasks, including Kaplan-Meier survival analysis in oncology and genome-wide association studies in medical genetics. Using our system, we accurately and efficiently reproduce two published centralized studies in a federated setting, enabling biomedical insights that are not possible from individual institutions alone. Our work represents a necessary key step towards overcoming the privacy hurdle in enabling multi-centric scientific collaborations.

bioinformatics↗

bioRxiv2021

Using population-specific add-on polymorphisms to improve genotype imputation in underrepresented populations

Genome-wide association studies rely on the statistical inference of untyped variants, called imputation, to increase the coverage of genotyping arrays. However, the results are often suboptimal in populations underrepresented in existing reference panels and array designs, since the selected single nucleotide polymorphisms (SNPs) may fail to capture population-specific haplotype structures, hence the full extent of common genetic variation. Here, we propose to sequence the full genome of a small subset of an underrepresented study cohort to inform the selection of population-specific add-on SNPs, such that the remaining array-genotyped cohort could be more accurately imputed. Using a Tanzania-based cohort as a proof-of-concept, we demonstrate the validity of our approach by showing improvements in imputation accuracy after the addition of our designed addon SNPs to the base H3Africa array.

genetics↗

bioRxiv2020

Ongoing evolution of KRAB zinc finger protein-coding genes in modern humans

BackgroundKruppel-associated box (KRAB) zinc finger proteins (KZFPs) constitute the largest and fastest evolving family of gene regulators encoded by the human genome. Recent data indicate that many KZFPs serve as repressors of transposable element-embedded regulatory sequences (TEeRS) and that the evolutionary turnover of KZFP genes is mainly attributable to the changing transposable element (TE) load of their hosts. However, how natural selection and genetic variation are shaping this process is still poorly defined. MethodsGenetic information was collected from nine primate species and 138,500 human genomes. Gene-wide as well as functional amino acid position specific constraint was calculated across all human KZFPs. ResultsWe found that the most conserved KZFPs, some of which go back close to 400 million years, have been subjected to marked negative selection in the evolutionarily recent past and are very homogeneous within the human population. In contrast, younger, largely primate-restricted family members present evidence of less negative selection than the rest of genome and lower levels of coding constraint, particularly within the sequences encoding the functional sites of their zinc finger (ZF) arrays. We defined 33 sets of KZFP paralogs, which pairwise displayed a broad range of coding constraints differentials, with more recently emerged paralogs usually displaying a higher frequency of putatively deleterious mutations and missense variants within the functional sites of their ZF arrays than their source gene. Finally, we identified three KZFP genes more constrained in the genomes of individuals of African ancestry than in Europeans, with their modes of expression or DNA targets pointing to possible links between these inter-populational genetic differences and regional differences in the prevalence of some diseases. ConclusionsThis work shows how the ongoing selection of KZFPs contributes to modern human genetic variation, in particular through the constraint of putatively deleterious- and missense variants in functional protein sites, and how ongoing interplays between environment and KZFP genes might be impacting the biology of modern humans.

genomics↗

bioRxiv2020

Phenotype prediction from genome-wide genotyping data: a crowdsourcing experiment

BackgroundThe increasing statistical power of genome-wide association studies is fostering the development of precision medicine through genomic predictions of complex traits. Nevertheless, it has been shown that the results remain relatively modest. A reason might be the nature of the methods typically used to construct genomic predictions. Recent machine learning techniques have properties that could help to capture the architecture of complex traits better and improve genomic prediction accuracy. MethodsWe relied on crowd-sourcing to efficiently compare multiple genomic prediction methods. This represents an innovative approach in the genomic field because of the privacy concerns linked to human genetic data. There are two crowd-sourcing elements building our study. First, we constructed a dataset from openSNP (opensnp.org), an open repository where people voluntarily share their genotyping data and phenotypic information in an effort to participate in open science. To leverage this resource we release the openSNP Cohort Maker, a tool that builds a homogeneous and up-to-date cohort based on the data available on opensnp.org. Second, we organized an open online challenge on the CrowdAI platform (crowdai.org) aiming at predicting height from genome-wide genotyping data. ResultsThe openSNP Height Prediction challenge lasted for three months. A total of 138 challengers contributed to 1275 submissions. The winner computed a polygenic risk score using the publicly available summary statistics of the GIANT study to achieve the best result (r2 = 0.53 versus r2 = 0.49 for the second-best). ConclusionWe report here the first crowd-sourced challenge on publicly available genome-wide genotyping data. We also deliver the openSNP Cohort Maker that will allow people to make use of the data available on opensnp.org.

genomics↗

bioRxiv2020

Viral genome wide association study identifies novel hepatitis C virus polymorphisms associated with sofosbuvir treatment failure

Chronic hepatitis C virus (HCV) infection is a major cause of chronic liver disease, cirrhosis and hepatocellular carcinoma worldwide. With the recent development of direct acting antivirals (DAA), treatment of chronically infected patients has become highly effective although a subset of patients do not respond to therapy. Sofosbuvir is a common component of current de novo or salvage combination therapies. We used pre-treatment whole genome sequencing of HCV from 507 patients infected with HCV subtype 3a and treated with sofosbuvir containing regimens to detect viral polymorphisms associated with response to treatment. We found that three common polymorphisms present in HCV NS2 and NS3 proteins (not direct targets of sofosbuvir) were associated with reduced treatment response. These polymorphisms were enriched in post-treatment HCV sequences of patients unresponsive to treatment; they were also associated with lower reductions in viral load in the first week of therapy. The finding of polymorphisms in NS2 and NS3 proteins associated with poor treatment outcomes emphasises the value of more systematic genome-wide analyses of HCV in uncovering indirect but clinically relevant mechanisms of antiviral resistance.

microbiology↗

bioRxiv2019

Genomics of Human Respiratory Syncytial Virus Vaccine Attenuation

The human orthopneumovirus (HRSV) is a major cause of lower respiratory tract infection in children worldwide. Despite decades of efforts, no vaccine is available. In this work, we report mutations that are frequent in vaccine candidates and rare in wild-type genomes, taking into account all the publicly available HRSV sequence data. These mutations are different from the ones already known to attenuate the virus, and thus may contribute to the effort towards producing a live attenuated vaccine against HRSV.

genomics↗

bioRxiv2019

Citizen-Centered, Auditable, and Privacy-Preserving Population Genomics

The growing number of health-data breaches, the use of genomic databases for law enforcement purposes and the lack of transparency of personal-genomics companies are raising unprecedented privacy concerns. To enable a secure exploration of genomic datasets with controlled and transparent data access, we propose a novel approach that combines cryptographic privacy-preserving technologies, such as homomorphic encryption and secure multi-party computation, with the auditability of blockchains. This approach provides strong security guarantees against realistic threat models by empowering individual citizens to decide who can query and access their genomic data and by ensuring end-to-end data confidentiality. Our open-source implementation supports queries on the encrypted genomic data of hundreds of thousands of individuals, with minimal overhead. Our work opens a path towards multi-functional, privacy-preserving genomic-data analysis.\n\nOne Sentence SummaryA citizen-centered open-source response to the privacy concerns that hinder population genomics, based on modern cryptography.

genomics↗

bioRxiv2019

A Comprehensive Assessment of Demographic, Environmental and Host Genetic Associations with Gut Microbiome Diversity in Healthy Individuals

BackgroundThe gut microbiome is an important determinant of human health. Its composition has been shown to be influenced by multiple environmental factors and likely by host genetic variation. In the framework of the Milieu Interieur Consortium, a total of 1,000 healthy individuals of western European ancestry, with a 1:1 sex ratio and evenly stratified across five decades of life (age 20 - 69), were recruited. We generated 16S ribosomal RNA profiles from stool samples for 858 participants. We investigated genetic and non-genetic factors that contribute to individual differences in fecal microbiome composition. ResultsAmong 110 demographic, clinical and environmental factors, 11 were identified as significantly correlated with -diversity, {beta}-diversity or abundance of specific microbial communities in multivariable models. Age and blood alanine aminotransferase levels showed the strongest associations with microbiome diversity. In total, all non-genetic factors explained 16.4% of the variance. We then searched for associations between >5 million single nucleotide polymorphisms and the same indicators of fecal microbiome diversity, including the significant non-genetic factors as covariates. No genome-wide significant associations were identified after correction for multiple testing. A small fraction of previously reported associations between human genetic variants and specific taxa could be replicated in our cohort, while no replication was observed for any of the diversity metrics. ConclusionIn a well-characterized cohort of healthy individuals, we identified several non-genetic variables associated with fecal microbiome diversity. In contrast, host genetics only had a negligible influence. Demographic and environmental factors are thus the main contributors to fecal microbiome composition in healthy individuals.

genomics↗

bioRxiv2018

Pervasive adaptation of hepatitis C virus to interferon lambda polymorphism across multiple genotypes

Genetic polymorphism in the interferon lambda (IFN-{lambda}) region is associated with spontaneous clearance of hepatitis C virus (HCV) infection and with response to interferon-based antiviral treatment. Here, we evaluate the associations between IFN- {lambda} polymorphism and HCV variation through a genome-to-genome analysis in 8,729 patients from diverse ancestral backgrounds infected with various HCV genotypes. We searched for associations between rs12979860 genotype, a tag for IFN-{lambda} haplotypes, and amino acid variants in the NS3, NS4A, NS5A and NS5B HCV proteins. We report multiple associations between host and pathogen variants in the full cohort as well as in subgroups defined by viral genotype and human ancestry. We also assess the combined impact of human and HCV variation on pre-treatment viral load. By demonstrating that IFN-{lambda} genetic variation leaves a large footprint in the viral genome, this study provides strong evidence of pervasive viral adaptation to host innate immune pressure during chronic HCV infection.

genomics↗

bioRxiv2018

Human genomics of acute liver failure due to hepatitis B virus infection: an exome sequencing study in liver transplant recipients

Acute liver failure (ALF) or fulminant hepatitis is a rare, yet severe outcome of infection with hepatitis B virus (HBV) that carries a high mortality rate. The occurrence of a life-threatening condition upon infection with a prevalent virus in individuals without known risk factors is suggestive of pathogen-specific immune dysregulation. In the absence of established differences in HBV virulence, we hypothesized that ALF upon primary infection with HBV could be due to rare deleterious variants in the human genome. To search for such variants, we performed exome sequencing in 21 previously healthy adults who required liver transplantation upon fulminant HBV infection and 172 controls that were positive for anti-HBc and anti-HBs antibodies but had no clinical history of jaundice or liver disease. After a series of hypothesis-driven filtering steps, we searched for putatively pathogenic variants that were significantly associated with case-control status. We did not find any causal variant or gene, a result that does not support the hypothesis of a shared monogenic basis for human susceptibility to HBV-related ALF in adults. This study represents a first attempt at deciphering the human genetic contribution to the most severe clinical presentation of acute HBV infection in previously healthy individuals.\n\nAuthor SummaryInfection with hepatitis B virus (HBV) is very common and causes a variety of liver diseases including acute and chronic hepatitis, cirrhosis and liver carcinoma. Acute HBV infection is often asymptomatic, still about 1% of newly infected people develop a rapid and severe disease known as acute liver failure or fulminant hepatitis. Acute liver failure has a high mortality rate and is an indication for urgent liver transplantation. It is not clear why some people, who are otherwise healthy, develop such severe symptoms upon infection with a common pathogen. Here, we hypothesized that rare DNA variants in the human genome could contribute to this unusual susceptibility. We sequenced the exome (i.e. the regions of the genome that encode the proteins) of 21 previously healthy adults who required liver transplantation upon fulminant HBV infection and searched for rare genetic variants that could explain the clinical presentation. We did not identify any variant that could be convincingly linked to the extreme susceptibility to HBV observed in the study participants. This suggests that HBV-induced acute liver failure is more likely to result from the combined influence of multiple genetic and environmental factors.

genomics↗

bioRxiv2018

HIV Peptidome-Wide Association Study Reveals Patient-Specific Epitope Repertoires Associated with HIV Control

Genetic variation in the peptide-binding groove of the highly polymorphic human leukocyte antigen (HLA) class I molecules has repeatedly been associated with HIV-1 control and progression to AIDS, accounting for up to 12% of the variation in HIV-1 set point viral load (spVL). This suggests a key role in disease control for HLA presentation of HIV-1 epitopes to cytotoxic T cells. However, a comprehensive understanding of the relevant HLA-bound HIV epitopes is still elusive. Here we developed a peptidome-wide association study (PepWAS) approach that integrates HLA genotypes and spVL data from 6,311 HIV-infected patients to interrogate the entire HIV-1 proteome (3,252 unique peptides) for disease-relevant peptides. This PepWAS approach revealed a core set of epitopes associated with spVL, including previously characterized epitopes but also several novel disease-relevant peptides. More importantly, each patient presents only 16 ({+/-}7) and 6 ({+/-}6) of these core epitopes through their individual HLA-B and HLA-A variants, respectively. Differences in these patient-specific epitope repertoires account for almost all the variation in spVL previously associated with HLA genetic variation. PepWAS thus enables a comprehensive functional interpretation of the robust but little understood association between HLA and HIV-1 control, prioritizing a short and targetable list of disease-associated epitopes for personalized immunotherapy.

genomics↗

bioRxiv2018

Sequence Alignment Through the Looking Glass

Rapid advances in sequencing technologies are producing genomic data on an unprecedented scale. The first, and often one of the most time consuming, step of genomic data analysis is sequence alignment, where sequenced reads must be aligned to a reference genome. Several years of research on alignment algorithms has led to the development of several state-of-the-art sequence aligners that can map tens of thousands of reads per second.\n\nIn this work, we answer the question \"How do sequence aligners utilize modern processors?\" We examine four state-of-the-art aligners running on an Intel processor and identify that all aligners leave the processor substantially underutilized. We perform an in-depth microarchitectural analysis to explore the interaction between aligner software and processor hardware. We identify bottlenecks that lead to processor underutilization and discuss the implications of our analysis on next-generation sequence aligner design.

bioinformatics↗

bioRxiv2018

Genetic factors and age are the strongest predictors of humoral immune responses to common pathogens and vaccines

IntroductionHumoral immune responses to infectious agents or vaccination vary substantially among individuals, and many of the factors responsible for this variability remain to be defined. Current evidence suggests that human genetic variation influences (i) serum immunoglobulin levels, (ii) seroconversion rates, and (iii) intensity of antigen-specific immune responses. Here, we evaluate the impact of intrinsic (age and sex), environmental and genetic factors on the variability of humoral response to common pathogens and vaccines.\n\nMethodsWe characterized the serological response to 15 antigens from common human pathogens or vaccines, in an age- and sex-stratified cohort of 1,000 healthy individuals (Milieu Interieur cohort). Using clinical-grade serological assays, we measured total IgA, IgE, IgG and IgM levels, as well as qualitative (serostatus) and quantitative IgG responses to cytomegalovirus, Epstein-Barr virus, herpes simplex virus 1 & 2, varicella zoster virus, Helicobacter pylori, Toxoplasma gondii, influenza A virus, measles, mumps, rubella, and hepatitis B virus. Following genome-wide genotyping of single nucleotide polymorphisms and imputation, we examined associations between ~5 million genetic variants and antibody responses using single marker and gene burden tests.\n\nResults and discussionWe identified age and sex as important determinants of humoral response, with older individuals and women having higher rates of seropositivity for most antigens. Genome-wide association studies revealed significant associations between variants in the human leucocyte antigen (HLA) class II region on chromosome 6 and anti-EBV and anti-rubella IgG levels. We used HLA imputation to fine map these associations to amino acid variants in the peptide-binding groove of HLA-DR{beta}1 and HLA-DP{beta}1, respectively. We also observed significant associations for total IgA levels with two loci on chromosome 2 and with specific KIR-HLA combinations.\n\nConclusionsUsing extensive serological testing and genome-wide association analyses in a well-characterized cohort of healthy individuals, we demonstrate that age, sex and specific human genetic variants contribute to inter-individual variability in humoral response. By highlighting genes and pathways implicated in the normal antibody response to frequently encountered antigens, these findings provide a basis to better understand disease pathogenesis.

genomics↗

bioRxiv2017

Correcting for population stratification reduces false positive and false negative results in joint analyses of host and pathogen genomes

Studies of host genetic determinants of pathogen sequence variation can identify sites of genomic conflicts, by highlighting variants that are implicated in immune response on the host side and adaptive escape on the pathogen side. However, systematic genetic differences in host and pathogen populations can lead to inflated type I (false positive) and type II (false negative) error rates in genome-wide association analyses. Here, we demonstrate through simulation that correcting for both host and pathogen stratification reduces spurious signals and increases power to detect real associations in a variety of tested scenarios. We confirm the validity of the simulations by showing comparable results in an analysis of paired human and HIV genomes.

genomics↗

bioRxiv2017

A genome-wide polygenic approach to HIV uncovers link to inflammatory bowel disease and identifies potential novel genetic variants

Polygenic approaches using genome-wide data have been hugely successful in confirming and quantifying the heritability of complex human traits. Here, we highlight their ability to identify potential novel risk variants by looking for variants with pleiotropic effect in genetically overlapping phenotypes.\n\nWe used LD Score Regression in a sample of 6,315 HIV+ European individuals and 7,247 controls to test for phenotypes genetically overlapping with susceptibility to HIV-1 infection. Using LD Hub, a web tool that performs LD Score Regression, identified two phenotypes with significant genetic overlap: schizophrenia (rG =0.19, p=0.0007 and ulcerative colitis (rG=0.22, p= 0.0061). We further showed that the genetic overlap between HIV acquisition and schizophrenia is likely driven in part by their shared overlap with cannabis use and sexual behavior. BUMHBOX analyses suggested that these genetic overlaps were driven by genome-wide pleiotropy with HIV acquisition rather than heterogeneity within the HIV acquisition sample. The two diseases identified as genetically overlapping with HIV-1 acquisition have >100 associated variants, and we tested if any of them significantly associated with HIV acquisition. We observed three variants that exceeded our threshold for statistical significance. Two of these were eQTLs in whole blood for genes coding for proteins suspected to be involved in HIV biology: rs1819333 in CCR6 (p=0.0002) and rs4932178 in FURIN (p=0.00033). However, no signal was found for these variants in two smaller African samples totaling 1015 cases and 963 controls, though the mode of acquisition and genetic architecture of these populations differed.\n\nThese results highlight the ability to use polygenic methods to gain new insights into complex diseases and identify potential associations with individual variants. Crucially, the leveraging of existing, publically available data makes these methods a cost-effective approach. In this case, our results add to the evidence for the role of risk taking behavior and inflammation of the bowel in HIV acquisition.\n\nAuthor SummaryThe biology of what puts certain individuals at greater risk of HIV acquisition is poorly understood. Using several novel polygenic methods, we identify supporting evidence for two important factors leading to acquisition. First, the role of an individuals genetic predisposition to risk taking behaviours such as number of sexual partners, age at first sexual intercourse drug use, and mental health problems. Second, the role of gut inflammation, in particular a genetic overlap between HIV acquisition with inflammatory bowel disease and the potential role of CCR6 during infection.

genomics↗

bioRxiv2017

Dissecting HIV Virulence: Heritability Of Setpoint Viral Load, CD4+ T Cell Decline And Per-Parasite Pathogenicity

Pathogen strains may differ in virulence because they attain different loads in their hosts, or because they induce different disease-causing mechanisms independent of their load. In evolutionary ecology, the latter is referred to as \"per-parasite pathogenicity\". Using viral load and CD4+ T cell measures from 2014 HIV-1 subtype B infected individuals enrolled in the Swiss HIV Cohort Study, we investigated if virulence -- measured as the rate of decline of CD4+ T cells -- and per-parasite pathogenicity are heritable from donor to recipient. We estimated heritability by donor-recipient regressions applied to 196 previously identified transmission pairs, and by phylogenetic mixed models applied to a phylogenetic tree inferred from HIV pol sequences. Regressing the CD4+ T cell declines and per-parasite pathogenicities of the transmission pairs did not yield heritability estimates significantly different from zero. With the phylogenetic mixed model, however, our best estimate for the heritability of the CD4+ T cell decline is 17% (5%-30%), and that of the per-parasite pathogenicity is 17% (4%-29%). Further, we confirm that the set-point viral load is heritable, and estimate a heritability of 29% (12%-46%). Interestingly, the pattern of evolution of all these traits differs significantly from neutrality, and is most consistent with stabilizing selection for the set-point viral load, and with directional selection for the CD4+ T cell decline and the per-parasite pathogenicity. Our analysis shows that the viral genetype affects virulence mainly by modulating the per-parasite pathogenicity, while the indirect effect via the set-point viral load is minor.

evolutionary biology↗

bioRxiv2016

Easy and Accurate Reconstruction of Whole HIV Genomes from Short-Read Sequence Data

Next-generation sequencing has yet to be widely adopted for HIV. The difficulty of accurately reconstructing the consensus sequence of a quasispecies from reads (short fragments of DNA) in the presence of rapid between- and within-host evolution may have presented a barrier. In particular, mapping (aligning) reads to a reference sequence leads to biased loss of information; this bias can distort epidemiological and evolutionary conclusions. De novo assembly avoids this bias by effectively aligning the reads to themselves, producing a set of sequences called contigs. However contigs provide only a partial summary of the reads, misassembly may result in their having an incorrect structure, and no information is available at parts of the genome where contigs could not be assembled. To address these problems we developed the tool shiver to preprocess reads for quality and contamination, then map them to a reference tailored to the sample using corrected contigs supplemented with existing reference sequences. Run with two commands per sample, it can easily be used for large heterogeneous data sets. We use shiver to reconstruct the consensus sequence and minority variant information from paired-end short-read data produced with the Illumina platform, for 65 existing publicly available samples and 50 new samples. We show the systematic superiority of mapping to shivers constructed reference over mapping the same reads to the standard reference HXB2: an average of 29 bases per sample are called differently, of which 98.5% are supported by higher coverage. We also provide a practical guide to working with imperfect contigs.

bioinformatics↗