Biology subjects

Oliver, C.

Publications and source records attributed to Oliver, C..

9 recordsLinked to original sources

bioRxiv2026

Adaptive 2.5D base-pairing subgraph search detects RNA small-molecule binding sites

Ribo-LENS is a geometric deep-learning framework for detecting small-molecule binding sites in RNA structures. It is designed to exploit two properties of RNA base-pairing networks: their robustness to conformational fluctuation and the functional signatures they encode. By reasoning directly in the space of base-pairing subgraphs, Ribo-LENS assembles coherent binding sites, in contrast to methods that score residues independently. In extensive experiments, Ribo-LENS is competitive with, and often outperforms, large all-atom co-folding models (AlphaFold3, Chai-1), fine-tuned language models (GerNA-Bind), and structure-based tools (RNAsite), raising mean MCC to 0.380 (versus 0.321 for the state-of-the-art GerNA-Bind). It is strongly robust to apo/holo rearrangement, with its accuracy tracking the base-pairing graph (Spearman{rho} = 0.82 with binding-site graph edit distance) rather than backbone displacement ({rho} = -0.15 with RMSD), and depends far less on sequence homology than competing predictors. In an end-to-end, sequence-based virtual screen of the ROBIN assay ([~]25,000 compounds), Ribo-LENS guides docking to a small predicted pocket, matching a blind all-atom cavity search (enrichment factors up to 6.1) at a fraction of the search cost; on two SARS-CoV-2 targets its predicted sites align with NMR chemical-shift perturbations. Ribo-LENS turns coarse base-pairing structure into a practical entry point for screening the vast, largely unexplored RNA target space.

bioinformatics↗

bioRxiv2025

Higher-Order Biomarkers Through Network Motif Mining: A COVID-19 Case Study

We introduce a novel approach for analyzing expression data by integrating patient-level expression profiles with a Protein-Protein interaction network from the STRING database. Our pipeline leverages motif mining to identify recurring sets (motifs) of interacting biomolecules characterized by specific expression patterns, providing deeper insights into underlying biological processes. We applied our method to a publicly available dataset of plasma protein measurements from patients with mild/moderate COVID-19 and compared the motif features to those found by conventional differential expression analysis. Motif features demonstrated better performance in classification models and hierarchical clustering. Of note, they were able to resolve interpatient variability during clustering, while traditional features failed to do so. Interestingly, these discriminatory performances were achieved using a smaller and largely different set of proteins. Motif mining is a highly flexible method with capacity to integrate multiple modes of data and presents an exciting line of analysis for biomarker discovery as well as general biology.

bioinformatics↗

bioRxiv2024

Gut MicrobiotAware: how much do we know about gut microbiota? An international questionnaire.

Over the past two decades, our understanding of human gut microbiota composition and function has grown significantly. The surge in studies and data has unveiled the profound influence of gut microbiota on host nutrition, health, and behavior, bridging biology, medicine, and ecology. The dynamic interaction between daily choices, life events, and gut microbiota composition makes it a fascinating subject for scientists to communicate to diverse audiences. Effective scientific communication involves adapting the delivery of knowledge to different audiences, using precise language in academic settings and accessible concepts in public forums like science festivals or social media. With the growing interest in gut microbiota beyond academic circles, there is an increased risk of disseminating information lacking scientific rigor. This study aims to assess the general publics knowledge about gut microbiota and evaluate healthcare professionals understanding of its links to various health conditions, ultimately informing better communication strategies for both groups.

scientific communication and education↗

bioRxiv2024

Local FK506 delivery induces osteogenesis in in vivo rat bone defect and rabbit spine fusion models

Bone grafting procedures are commonly used for the repair, regeneration, and fusion of bones in in a wide range of orthopaedic surgeries, including large bone defects and spine fusion procedures. Autografts are the clinical gold standard, though recombinant human bone morphogenetic proteins (rhBMPs) are often used, particularly in difficult clinical situations. However, treatment with rhBMPs can have off-target effects and significantly increase surgical costs, adding to patients already high economic and mental burden. Recent studies have identified that FDA-approved immunosuppressant drug, FK506 (Tacrolimus), can also activate the BMP pathway by binding to its inhibitors. This study tested the hypothesis that FK506, as a standalone treatment, could induce osteogenic differentiation of human mesenchymal stromal cells (hMSCs), as well as functional bone formation in a rat segmental bone defect model and rabbit spinal fusion model. FK506 potentiated the effect of low dose BMP-2 to enhance osteogenic differentiation and mineralization of hMSCs in vitro. Standalone treatment with FK506 delivered on a collagen sponge, produced consistent bone bridging of a rat critically-sized femoral defect with functional mechanical properties comparable to naive bone. In a rabbit single level posterolateral spine fusion model, treatment with FK506 delivered on a collagen sponge successfully fused the L5-L6 vertebrae at rates comparable to rhBMP-2 treatment. These data demonstrate the ability of FK506 to induce bone formation in human cells and two challenging in vivo models, and indicate FK506 can be utilized either as a standalone treatment or in conjunction with rhBMP to treat a variety of spine disorders. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=82 SRC="FIGDIR/small/584163v1_ufig1.gif" ALT="Figure 1"> View larger version (29K): org.highwire.dtl.DTLVardef@1b0b690org.highwire.dtl.DTLVardef@a1e032org.highwire.dtl.DTLVardef@1862a55org.highwire.dtl.DTLVardef@1456fc1_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioengineering↗

bioRxiv2023

Semi-supervised learning and large-scale docking data accelerate RNA virtual screening

RNAs constitute a vast reservoir of mostly untapped drug targets. Structure-based virtual screening (VS) methods screen large compound libraries for identifying promising candidate molecules by conditioning on binding site information. The classical approach relies on molecular docking simulations. However, this strategy does not scale well with the size of the small molecule databases and the number of potential RNA targets. Machine learning emerged as a promising technology to resolve this bottleneck. Efficient data-driven VS methods have already been introduced for proteins, but these techniques have not yet been developed for RNAs due to limited dataset sizes and lack of practical use-case evaluation. We propose a data-driven VS pipeline that deals with the unique challenges of RNA molecules through coarse grained modeling of 3D structures and heterogeneous training regimes using synthetic data augmentation and RNA-centric self supervision. We report strong prediction and generalizability of our framework, ranking active compounds among inactives in the top 2.8% on average on a structurally distinct drug-like test set. Those predictions are sensitive, but robust to pockets alterations, opening the door to its use on binding site detection methods outputs. Our model results in a ten thousand-times speedup over docking techniques while obtaining higher performance. Finally, we deploy our model on a recently published in-vitro small molecule microarray experiment with 20,000 compounds and report a mean enrichment factor at 1% of 2.93 on four unseen RNA riboswitches. To our knowledge, this is the first experimental evidence of success for structure-based deep learning methods in RNA virtual screening. Our source code and data, as well as a Google Colab notebook for inference, are available on GitHub.1

bioinformatics↗

bioRxiv2021

Expression dynamics of ARGONAUTE proteins during meiosis in Arabidopsis

Meiosis is a specialized cell division that is key for reproduction and genetic diversity in sexually reproducing plants. Recently, different RNA silencing pathways have been proposed to carry a specific activity during meiosis, but the pathways involved during this process remain unclear. Here, we explored the subcellular localization of different ARGONAUTE (AGO) proteins, the main effectors of RNA silencing, during male meiosis in Arabidopsis thaliana using immunolocalizations with commercially available antibodies. We detected the presence of AGO proteins associated with posttranscriptional gene silencing (AGO1, 2 and 5) in the cytoplasm or the nucleus, while AGOs associated with transcriptional gene silencing (AGO4 and 9) localized exclusively in the nucleus. These results indicate that the localization of different AGOs correlates with their predicted roles at the transcriptional and posttranscriptional levels and provide an overview of their timing and potential role during meiosis.

plant biology↗

bioRxiv2020

OptiMol: Optimization of binding affinities in chemical space for drug discovery

Ligand-based drug design has recently benefited from the boost of deep generative models. These models enable extensive explorations of the chemical space, and provide a platform for molecular optimization. However, current state of the art methods do not leverage the structure of the target, which is known to play a key role in the interaction. We propose an optimization pipeline that leverages complementary structure-based and ligand-based methods. Instead of performing docking on a fixed drug bank, we iteratively select promising compounds in the whole chemical space using a ligand-centered generative model. Molecular docking is then used as an oracle to guide compound optimization. This allows to iteratively generate leads that better fit the target structure, in a closed optimization loop, without prior knowledge about bio-actives. For this purpose, we introduce a new graph to selfies VAE which benefits from a seventeen times faster decoding than graph to graph methods while being competitive with the state of the art. We then successfully optimize the generation of molecules towards high docking scores, enabling a ten-fold augmentation of high-scoring compounds found with a fixed computational budget. AvailabilityCode is available on GitHub Contactjeromew@mcgill.ca

bioinformatics↗

bioRxiv2019

MicroRNA function transitions from regulating developmental genes to transposable elements during the maturation of pollen

microRNAs play important roles to control the development of eukaryotic organisms. Both animal and plant microRNAs are essential for the spatio-temporal regulation of development but together with this role, plant microRNAs also control transposable elements and stimulate the production of epigenetically-active small interfering RNAs. This last role is evident in the plant male gamete containing structure, the male gametophyte or pollen grain, but how the dual role of plant microRNAs is integrated during its development is unknown. Here, we provide a detailed analysis of microRNA dynamics during pollen development and their genic and transposable element targets using small RNA and mRNA cleavage (PARE) high-throughput sequencing. Furthermore we uncover the microRNAs loaded in the two main Argonaute proteins in the mature pollen grain, AGO1 and AGO5. Our results indicate that the developmental progression from microspore to mature pollen grain is characterized by a reprogramming from microRNAs focused on the control of development to microRNAs regulating transposable element control.

plant biology↗

bioRxiv2018

Mouse REC114 is essential for meiotic DNA double-strand break formation and forms a complex with MEI4

Programmed formation of DNA double strand breaks (DSBs) initiates the meiotic homologous recombination pathway. This pathway allows homologous chromosomes to find each other and the formation of crossing overs, the products of reciprocal exchanges, which are required for proper chromosome segregation at the first meiotic division. Meiotic DSBs are catalyzed by Spo11 that forms a complex with a second subunit, TopoVIBL, and mediates a DNA type II topoisomerase-like cleavage. Several other proteins are essential for meiotic DSB formation, including three evolutionarily conserved proteins first identified in Saccharomyces cerevisiae (Mer2, Mei4 and Rec114). These three S. cerevisiae proteins and their mouse orthologs (IHO1, MEI4 and REC114) co-localize on the axes of meiotic chromosomes, and mouse IHO1 and MEI4 are essential for meiotic DSB formation. Here, we show that mouse Rec114 is required for meiotic DSB formation. Moreover, MEI4 forms a complex with REC114 and IHO1 in mouse spermatocytes, consistent with cytological observations. We then demonstrated in vitro the formation of a stable complex between REC114 C-terminal domain and MEI4 N-terminal domain. We further determine the structure of REC114 N-terminal domain that revealed similarity with Pleckstrin Homology domains and its property to dimerize. These analyses provide direct insights into the architecture of these essential components of the meiotic DSB machinery.

molecular biology↗