Biology subjects

Mourad, R.

Publications and source records attributed to Mourad, R..

12 recordsLinked to original sources

bioRxiv2026

SPA-C: an hybrid tool to accurately scaffold genomes using Hi-C and Deep-Learning

Genome assembly is a computational pipeline designed to reconstruct chromosomes from small sequencing reads. Following their assembly, contiguous sequences (contigs) are arranged into chromosome-long sequences during scaffolding. Hi-C, a long-range linkage information between regions of the genome widely used in recent large sequencing projects, is often required to correctly order contigs. Several tools have been developed to automate this task following either statistical or deep-learning approaches. Statistical approaches summarise 2D Hi-C matrices into contact densities across sequences, thus ignoring informative visual patterns. The sole existing deep-learning tool uses a transformer-based computer vision model to correct the assembly. It has been trained on several species and uses Hi-C matrices directly. Yet it comes as a supplementary step in the scaffolding process, introducing extra computation time, and has been trained on a dataset that might contain labelling errors, which could provide sub-optimal results. We propose SPA-C, an hybrid pipeline combining the strengths of both approaches. Linkage prediction is handled with a frugal CNN-based model and a graph-solving algorithm is used to generate the scaffolds. Through our inputs design, the model is able to both correct errors within assemblies and link contigs, leveraging small, local Hi-C contact matrices. We handled low-complexity regions that might induce erroneous predictions using an external tool, improving the overall accuracy of generated assemblies. On a benchmark of six various genomes and four standard metrics, SPA-C outperformed four out of four state-of-the-art methods while achieving comparable start-to-end computation time. Python and Bash scripts are available on GitHub (github.com/SPA-C/SPA-C.git) and Zenodo (10.5281/zenodo.19000361).

bioinformatics↗

bioRxiv2025

Neural network-based cross-species chromatin annotation goes beyond sequence conservation

Analogous to the Encyclopedia of DNA Elements (ENCODE) project, the Functional Annotation of ANimal Genomes (FAANG) consortium has produced chromatin annotations for domesticated animals, albeit in smaller amounts. Although acquiring experimental data is more accessible and affordable for many species, human and mouse organisms will remain the reference. Classical methods based on sequence conservation can be used to infer missing annotations, but are inappropriate for non-conserved sequences. While regulatory sequences share low to moderate conservation, they have retained their regulatory function during the evolution process. Here, we take advantage of three neural networks (DeepBind, DeepSEA, and Enformer) trained with human and mouse ENCODE data to infer chromatin annotations (transcription factors binding, chromatin accessibility, and histone marks) in cattle, pig, chicken, and European seabass. For this purpose, we comprehensively assessed the quality of predictions using experimental data from FAANG, through AUC-ROC and AUC-PR metrics. Our results showed similar predictions for various annotations in mammals and chicken, with AUC-PR ranging from 0.157 to 0.765 for H3K4me1 and H3K4me3, respectively, but lower in fish. Further analyses focused on pigs highlighted (i) accurate predictions even for non-conserved sequences, and (ii) variable predictions depending on genomic feature annotations. Our results advocate the widespread use of human-trained neural networks as a first step in cross-species genome annotation before training species-specific models.

genomics↗

bioRxiv2023

Semi-supervised learning with pseudo-labeling for regulatory sequence prediction

Predicting molecular processes using deep learning is a promising approach to provide biological insights for non-coding SNPs identified in genome-wide association studies. However, most deep learning methods rely on supervised learning, which requires DNA sequences associated with functional data, and whose amount is severely limited by the finite size of the human genome. Conversely, the amount of mammalian DNA sequences is growing exponentially due to ongoing large-scale sequencing projects, but in most cases without functional data. To alleviate the limitations of supervised learning, we propose a novel semi-supervised learning based on pseudo-labeling, which allows to exploit unlabeled DNA sequences from numerous genomes during model pre-training. The approach is very flexible and can be used to train any neural architecture including state-of-the-art models, and shows in certain situations strong predictive performance improvements compared to standard supervised learning in most cases. Moreover, small models trained by SSL showed similar or better performance than large language model DNABERT2.

bioinformatics↗

bioRxiv2023

ProA and ProB repeat sequences shape genome organization, and enhancers open domains

Genome organization is partially conserved across cell types, yet its DNA-encoded determinants remain incompletely understood. Here we define ProA and ProB repeat sequences (RepSeqs) as two classes of cis-elements that promote A/euchromatin or B/heterochromatin compartment identity. We show that relative ProA/ProB density predicts Hi-C compartment profiles, indicating that compartmental propensity is largely encoded in sequence composition, and point to specific chromatin-based mechanisms underlying these effects. ProA RepSeqs are predominantly Alu elements, whereas ProB RepSeqs comprise young LINE-1s, selected ERVs, AT-rich microsatellites, and satellite repeats. RepSeqs of more indefinite character, including transcriptional enhancers, can switch between ProA and ProB functions to open or close chromatin domains in a context-dependent manner. In cancer, CpG methylation loss disproportionately impacts ProB RepSeqs, weakening the B compartment and thereby contributing to genome unfolding and cancer cell plasticity.

genetics↗

bioRxiv2023

G-quadruplexes are promoter elements controlling nucleosome exclusion and RNA polymerase II pausing

Despite their central role in transcription, it has been difficult to define universal sequences associated to eukaryotic promoters. Within chromatin context, recruitment of the transcriptional machinery requires opening of the promoter but how DNA elements could contribute to this process has remained elusive. Here, we show that G-quadruplex (G4) secondary structures are highly enriched mammalian core promoter elements. G4s are located at the deepest point of nucleosome exclusion at promoters and correlate with maximum promoter activity. We found that experimental G4s exclude nucleosomes both in vivo and in vitro and display a strong positioning potential. At model promoters, impairing G4s affected both transcriptional activity and chromatin opening. G4 destabilization also resulted in an inactive promoter state and affected transition to effective RNA production in live imaging experiments. Finally, G4 stabilization resulted in global reduction of proximal promoter pausing. Altogether, our data introduce G4s as bona fide promoter elements allowing nucleosome exclusion and facilitating pause release by the RNA Polymerase II.

genomics↗

bioRxiv2022

Semi-supervised deep learning with graph neural network for cross-species regulatory sequence prediction

Genome-wide association studies have systematically identified thousands of single nucleotide polymorphisms (SNPs) associated with complex genetic diseases. However, the majority of those SNPs were found in non-coding genomic regions, preventing the understanding of the underlying causal mechanism. Predicting molecular processes based on the DNA sequence represents a promising approach to understand the role of those non-coding SNPs. Over the past years, deep learning was successfully applied to regulatory sequence prediction. Such method required DNA sequences associated with functional data for training. However, the human genome has a finite size which strongly limits the amount of DNA sequence with functional data available for training. Conversely, the amount of mammalian DNA sequences is exponentially increasing due to ongoing large sequencing projects, but without functional data in most cases. Here, we propose a semi-supervised learning approach based on graph neural network which allows to borrow information from homologous mammal sequences during training. Our approach can be plugged into any existing deep learning model and showed improvements in many different situations, including classification and regression, and for different types of functional data.

bioinformatics↗

bioRxiv2021

ATM-dependent formation of a novel chromatin compartment regulates the Response to DNA Double Strand Breaks and the biogenesis of translocations

DNA Double-Strand Breaks (DSBs) repair is essential to safeguard genome integrity but the contribution of chromosome folding into this process remains elusive. Here we unveiled basic principles of chromosome dynamics upon DSBs in mammalian cells, controlled by key kinases from the DNA Damage Response. We report that ATM is responsible for the reinforcement of topologically associating domains (TAD) that experience a DSB. ATM further drives the formation of a new chromatin sub-compartment ("D" compartment) upon clustering of damaged TADs decorated with {gamma}H2AX and 53BP1. "D" compartment formation mostly occurs in G1, is independent of cohesin and is enhanced upon DNA-PK pharmacological inhibition. Importantly, a subset of DNA damage responsive genes that are upregulated following DSBs also physically localize in the D sub-compartment and this ensures their optimal activation, providing a function for DSB clustering in activating the DNA Damage Response. However, these DSB-induced changes in genome organization also come at the expense of an increased translocations rate, which we could also detect on cancer genomes. Overall, our work provides a function for DSB-induced compartmentalization in orchestrating the DNA Damage Response and highlights the critical impact of chromosome architecture in genomic instability.

molecular biology↗

bioRxiv2020

DeepG4 : A deep learning approach to predict active G-quadruplexes

DNA is a complex molecule carrying the instructions an organism needs to develop, live and reproduce. In 1953, Watson and Crick discovered that DNA is composed of two chains forming a double-helix. Later on, other structures of DNA were discovered and shown to play important roles in the cell, in particular G-quadruplex (G4). Following genome sequencing, several bioinformatic algorithms were developed to map G4s in vitro based on a canonical sequence motif, G-richness and G-skewness or alternatively sequence features including k-mers, and more recently machine/deep learning. Here, we propose a novel convolutional neural network (DeepG4) to map active G4s (forming both in vitro and in vivo). DeepG4 is very accurate to predict active G4s, while most state-of-the-art algorithms fail. Moreover, DeepG4 identifies key DNA motifs that are predictive of G4 activity. We found that active G4 motifs do not follow a very flexible sequence pattern as current algorithms seek for. Instead, active G4s are determined by numerous specific motifs. Moreover, among those motifs, we identified known transcription factors (TFs) which could play important roles in G4 activity by contributing either directly to G4 structures themselves or indirectly by participating in G4 formation in the vicinity. Moreover, we showed that specific TFs might explain G4 activity depending on cell type. Lastly, variant analysis suggests that SNPs altering predicted G4 activity could affect transcription and chromatin, e.g. gene expression, H3K4me3 mark and DNA methylation. Thus, DeepG4 paves the way for future studies assessing the impact of known disease-associated variants on DNA secondary structure by providing a mechanistic interpretation of SNP impact on transcription and chromatin. Availability: https://github.com/morphos30/DeepG4. Author summaryDNA is a molecule carrying genetic information and found in all living cells. In 1953, Watson and Crick found that DNA has a double helix structure. However, other DNA structures were later identified, and most notably, G-quadruplex (G4). In 2000, the Human Genome Project revealed the widespread presence of G4s in the genome using algorithms. To date, all G4 mapping algorithms were developed to map G4s on naked DNA, without knowing if they could be formed in the cell. Here, we designed a novel artificial intelligence algorithm that could map G4s active in the cell from the DNA sequence. We showed its better accuracy compared to existing algorithms. Moreover, we identified key transcriptional factor motifs that could explain G4 activity depending on cell type. Lastly, we demonstrated the existence of mutations that could alter G4 activity and therefore impact molecular processes, such as transcription, in the cell. Such results could provide a novel mechanistic interpretation of known disease-associated mutations.

bioinformatics↗

bioRxiv2020

BLM-dependent Break-Induced Replication handles DSBs in transcribed chromatin upon impaired RNA:DNA hybrids dissolution.

Transcriptionally active loci are particularly prone to breakage and mounting evidence suggest that DNA Double-Strand Breaks arising in genes are handled by a dedicated repair pathway, Transcription-Coupled DSB Repair (TC-DSBR), that entails R-loops accumulation and dissolution. Here, we uncovered a critical function of the Bloom RecQ DNA helicase (BLM) in TC-DSBR in human cells. BLM is recruited in a transcription dependent-manner at DSBs where it fosters resection, RAD51 binding and accurate Homologous Recombination repair. However, in a R-loop dissolution-deficient background BLM switches from promoting Homologous Recombination to promoting Break-Induced Replication (BIR), which strongly impairs cell viability. Altogether our work unveils a role for BLM in BIR at DSBs in active chromatin, and highlights the toxic potential of RNA:DNA hybrids that accumulate at these transcription-associated DSBs.

molecular biology↗

bioRxiv2020

Loop extrusion as a mechanism for DNA Double-Strand Breaks repair foci formation

DNA Double-Strand Breaks (DSBs) repair is essential to safeguard genome integrity. Upon DSBs, the ATM PI3K kinase rapidly triggers the establishment of megabase-sized, {gamma}H2AX-decorated chromatin domains which further act as seeds for the formation of DNA Damage Response (DDR) foci1. How these foci are rapidly assembled in order to establish a "repair-prone" environment within the nucleus is yet unclear. Topologically Associating Domains (TADs) are a key feature of 3D genome organization that regulate transcription and replication, but little is known about their contribution to DNA repair processes2,3. Here we found that TADs are functional units of the DDR, instrumental for the correct establishment of {gamma}H2AX/53BP1 chromatin domains in a manner that involves one-sided cohesin-mediated loop extrusion on both sides of the DSB. We propose a model whereby H2AX-containing nucleosomes are rapidly phosphorylated as they actively pass by DSB-anchored cohesin. Our work highlights the critical impact of chromosome conformation in the maintenance of genome integrity and provides the first example of a chromatin modification established by loop extrusion.

molecular biology↗

bioRxiv2019

Studying 3D genome evolution using genomic sequence

The 3D genome is essential to numerous key processes such as the regulation of gene expression and the replication-timing program. In vertebrates, chromatin looping is often mediated by CTCF, and marked by CTCF motif pairs in convergent orientation. Comparative Hi-C recently revealed that chromatin looping evolves across species. However, Hi-C experiments are complex and costly, which currently limits their use for evolutionary studies over a large number of species. Here, we propose a novel approach to study the 3D genome evolution in vertebrates using the genomic sequence only, e.g. without the need for Hi-C data. The approach is simple and relies on comparing the distances between convergent and divergent CTCF motifs (ratio R). We show that R is a powerful statistic to detect CTCF looping encoded in the human genome sequence, thus reflecting strong evolutionary constraints encoded in DNA and associated with the 3D genome. When comparing vertebrate genomes, our results reveal that R which underlies CTCF looping and TAD organization evolves over time and suggest that ancestral character reconstruction can be used to infer R in ancestral genomes.

bioinformatics↗

bioRxiv2017

Predicting double-strand DNA breaks using epigenome marks or DNA at kilobase resolution

Double-strand breaks (DSBs) result from the attack of both DNA strands by multiple sources, including exposure to ionizing radiation or reactive oxygen species. DSBs can cause abnormal chromosomal rearrangements which are linked to cancer development, and hence represent an important issue. Recent techniques allow the genome-wide mapping of DSBs at high resolution, enabling the comprehensive study of DSB origin. However these techniques are costly and challenging. Hence we devised a computational approach to predict DSBs using the epigenomic and chromatin context, for which public data are available from the ENCODE project. We achieved excellent prediction accuracy (AUC = 0.97) at high resolution (< 1 kb), and showed that only chromatin accessibility and H3K4me1 mark were sufficient for highly accurate prediction (AUC = 0.95). We also demonstrated the better sensitivity of DSB predictions compared to BLESS experiments. We identified chromatin accessibility, activity and long-range contacts as best predictors. In addition, our work represents the first step toward unveiling the\"cis-DNA repairing\" code underlying DSBs, paving the way for future studies of cis-elements involved in DNA damage and repair.

bioinformatics↗