Biology subjects

Ghoussaini, M.

Publications and source records attributed to Ghoussaini, M..

5 recordsLinked to original sources

bioRxiv2025

Variant Classification Using Proteomics-Informed Large Language Models Increases Power of Rare Variant Association Studies and Enhances Target Discovery

Rare variant association analysis, which assesses the aggregate effect of rare damaging variants within a gene, is a powerful strategy for advancing knowledge of human biology. Numerous models have been proposed to identify damaging coding variants, with the most recent ones employing deep learning and large language models (LLM) to predict the impact of changes in coding sequences. Here, we use newly available proteomics data on 2,898 proteins across 46, 665 individuals to evaluate and refine LLM predictors of damaging variants. Using one of these refined models, we evaluate association between rare damaging variants and human phenotypes at 241 positive control gene-trait pairs. Among these gene-trait pairs, our proteomics-guided model outperforms an ensemble of conventional approaches including PolyPhen2, Mutation Taster, SIFT, and LRT, as well as newer machine learning approaches for identifying damaging missense variants, such as CADD, ESM-1v, ESM-1b and AlphaMissense. When attempting to recover known associations by correctly separating damaging singleton missense variants from other singleton variants, our approach recapitulates 36.5% of gene-trait pairs with known associations, exceeding all the alternatives we considered. Furthermore, when we apply our model to 10 exemplary traits from the UK Biobank, we identify 177 gene-trait associations - again exceeding all other approaches. Our results demonstrate that summary statistics from large-scale human proteomics data enable evaluation and refinement of coding variant classification LLMs, improving discovery potential in human genetic studies.

bioinformatics↗

bioRxiv2021

Mapping interindividual dynamics of innate immune response at single-cell resolution

Common genetic variants modulate the cellular response to viruses and are implicated in a range of immune pathologies, including infectious and autoimmune diseases. The transcriptional antiviral response is known to vary between infected cells from a single individual, yet how genetic variants across individuals modulate the antiviral response (and its cell-to-cell variability) is not well understood. Here, we triggered the antiviral response in human fibroblasts from 68 healthy donors, and profiled tens of thousands of cells using single-cell RNA-seq. We developed GASPACHO (GAuSsian Processes for Association mapping leveraging Cell HeterOgeneity), the first statistical approach designed to identify dynamic eQTLs across a transcriptional trajectory of cell populations, without aggregating single-cell data into pseudo-bulk. This allows us to uncover the underlying architecture and variability of antiviral response across responding cells, and to identify more than two thousands eQTLs modulating the dynamic changes during this response. Many of these eQTLs colocalise with risk loci identified in GWAS of infectious and autoimmune diseases. As a case study, we focus on a COVID-19 susceptibility locus, colocalised with the antiviral OAS1 splicing QTL. We validated it in blood cells from a patient cohort and in the infected nasal cells of a patient with the risk allele, demonstrating the utility of GASPACHO to fine-map and functionally characterise a genetic locus. In summary, our novel analytical approach provides a new framework for delineation of the genetic variants that shape a wide spectrum of transcriptional responses at single-cell resolution.

genetics↗

bioRxiv2020

Open Targets Genetics: An open approach to systematically prioritize causal variants and genes at all published GWAS trait-associated loci

Genome-wide association studies (GWAS) have identified many variants robustly associated with complex traits but identifying the gene(s) mediating such associations is a major challenge. Here we present an open resource that provides systematic fine-mapping and protein-coding gene prioritization across 133,441 published human GWAS loci. We integrate diverse data sources, including genetics (from GWAS Catalog and UK Biobank) as well as transcriptomic, proteomic and epigenomic data across many tissues and cell types. We also provide systematic disease-disease and disease-molecular trait colocalization results across 92 cell types and tissues and identify 729 loci fine-mapped to a single coding causal variant and colocalized with a single gene. We trained a machine learning model using the fine mapped genetics and functional genomics data using 445 gold standard curated GWAS loci to distinguish causal genes from background genes at the same loci, outperforming a naive distance based model. Genes prioritized by our model are enriched for known approved drug targets (OR = 8.1, 95% CI: [5.7, 11.5]). These results will be regularly updated and are publicly available through a web portal, Open Targets Genetics (OTG, http://genetics.opentargets.org), enabling users to easily prioritize genes at disease-associated loci and assess their potential as drug targets.

genetics↗

bioRxiv2020

Population-scale single-cell RNA-seq profiling across dopaminergic neuron differentiation

Common genetic variants can have profound effects on cellular function, but studying these effects in primary human tissue samples and during development is challenging. Human induced pluripotent stem cell (iPSC) technology holds great promise for assessing these effects across different differentiation contexts. Here, we use an efficient pooling strategy to differentiate 215 iPS cell lines towards a midbrain neural fate, including dopaminergic neurons, and profile over 1 million cells sampled across three differentiation timepoints using single cell RNA sequencing. We find that the proportion of neuronal cells produced by each cell line is highly reproducible over different experimental batches, and identify robust molecular markers in pluripotent cells that predict line-to-line differences in cell fate. We identify expression quantitative trait loci (eQTL) that manifest at different stages of neuronal development, and in response to oxidative stress, by exposing cells to rotenone. We find over one thousand eQTL that colocalise with a known risk locus for a neurological trait, nearly half of which are not found in GTEx. Our study illustrates how coupling single cell transcriptomics with long-term iPSC differentiation can profile mechanistic effects of human trait-associated genetic variants in otherwise inaccessible cell states.

genetics↗

bioRxiv2019

Fine-mapping of 150 breast cancer risk regions identifies 178 high confidence target genes

Genome-wide association studies have identified breast cancer risk variants in over 150 genomic regions, but the mechanisms underlying risk remain largely unknown. These regions were explored by combining association analysis with in silico genomic feature annotations. We defined 205 independent risk-associated signals with the set of credible causal variants (CCVs) in each one. In parallel, we used a Bayesian approach (PAINTOR) that combines genetic association, linkage disequilibrium, and enriched genomic features to determine variants with high posterior probabilities (HPPs) of being causal. Potentially causal variants were significantly over-represented in active gene regulatory regions and transcription factor binding sites. We applied our INQUSIT pipeline for prioritizing genes as targets of potentially causal variants, using gene expression (eQTL), chromatin interaction and functional annotations. Known cancer drivers, transcription factors and genes in the developmental, apoptosis, immune system and DNA integrity checkpoint gene ontology pathways, were over-represented among the 178 highest confidence target genes.

genetics↗