Biology subjects

Stoven, V.

Publications and source records attributed to Stoven, V..

8 recordsLinked to original sources

bioRxiv2026

Transcriptomic profile of MTUS1-low TNBC reveals candidate therapeutic strategies.

BackgroundTriple-negative breast cancer (TNBC) is a clinically aggressive breast cancer subtype. It is a heterogeneous disease that remains difficult to stratify and that still lacks durable and biomarker-guided therapeutic options. Low expression of the tumour suppressor MTUS1 is associated with aggressive breast cancer features, but the biological properties of MTUS1-low TNBC remain insufficiently defined. Our goal was to determine whether low MTUS1 expression defines shared proliferative and stress-adaptation mechanisms that could guide candidate therapeutic strategies and corresponding target/drug pairs in MTUS1-low TNBC. MethodsWe labelled tumours from seven public TNBC RNA-seq cohorts based on the lowest and highest MTUS1 expression tertiles. Differential gene expression was analysed using gene set enrichment analysis (GSEA) on the Hallmark pathway database to identify deregulated biological pathways between MTUS1-low TNBC tumours and their MTUS1-high counterparts. Reproducibility was examined across independent TNBC cohorts and secondarily in broader breast cancer and selected TCGA tumour cohorts. Gene essentiality scores from CRISPR-Cas9 experiments in TNBC cell-line models were correlated to MTUS1 expression in these cell lines, to propose therapeutic strategies and their corresponding candidate target/drug pairs. ResultsMTUS1-low tumours showed a reproducible pathway-level proliferation mechanism driven by the MYC oncogene and sustained by up-regulated oxidative phosphorylation, combined with stress adaptation mechanisms involving unfolded protein response (UPR), and DNA repair Hallmark gene sets. Based on CRISPR data, we propose 3 therapeutic strategies: (1) targeting MYC to reduce its transcriptional activity, (2) targeting proteins from UPR, (3) targeting DNA-repair. We also propose corresponding candidate target/drug pairs to allow experimental validation of these strategies. ConclusionsProliferation in low MTUS1 TNBC is driven by MYC and stress-adaptation mechanisms. By linking this tumour profile to CRISPR-derived dependency signals, our analysis prioritises experimentally testable target-pathway hypotheses centred on MYC, UPR/proteostasis, and DNA-repair or checkpoint control. Although the proposed therapeutic strategies and candidate targets remain to be experimentally tested, the latter finding is consistent with published work showing that ATIP3-deficient TNBC cell line models are sensitive to inhibition of the WEE1 PKMYT1 G2/M checkpoint kinases.

cancer biology↗

bioRxiv2024

A molecular representation to identify isofunctional molecules

The challenges of drug discovery from hit identification to clinical development sometimes involve addressing scaffold hopping issues, in order to optimize biological activity or ADME properties, improve selectivity or mitigate toxicology concerns of a drug candidate, not to mention intellectual property reasons. Docking is usually viewed as the method of choice for identification of isofunctional molecules, i.e. highly dissimilar molecules that share common binding modes with a protein target. However, in cases where the protein structure has low resolution or is unknown, docking may not be suitable. In such cases, ligand-based approaches offer promise but are often inadequate to handle large-step scaffold hopping, because they usually rely on the molecular structure. Therefore, we propose the Interaction Fingerprints Profile (IFPP), a molecular representation that captures molecules binding modes based on docking experiments against a panel of diverse high-quality protein structures. Evaluation on the Large-Hops (LH) benchmark demonstrates the utility of IFPP for identification of isofunctional molecules. Nevertheless, computation of IFPPs is expensive, which limits the scalability for screening very large molecular libraries. We propose to overcome this limitation by leveraging Metric Learning approaches, allowing fast estimation of molecules IFPP similarities, thus providing an efficient pre-screening strategy applicable to very large molecular libraries. Overall, our results suggest that IFPP provides an interesting and complementary tool alongside existing methods, in order to address challenging scaffold hopping problems effectively in drug discovery.

bioinformatics↗

bioRxiv2024

Advancing Drug-Target Interactions Prediction: Leveraging a Large-Scale Dataset with a Rapid and Robust Chemogenomic Algorithm.

Drug-target interactions (DTIs) prediction algorithms are used are various stages of the drug discovery process. In this context, specific problems such as de-orphanization of a new therapeutic target, or target identification of a drug candidate arising from phenotypic screens require large-scale predictions across the protein and molecule spaces. DTI prediction heavily relies on supervised learning algorithms that use known DTIs to learn associations between molecule and protein features, allowing for the prediction of new interactions based on learned patterns. The algorithms must be broadly applicable to enable reliable predictions, even in regions of the protein or molecule spaces where data may be scarce. In this paper, we address two key challenges to fulfil these goals: building large, high-quality training datasets and designing prediction methods that can scale, in order to be trained on such large datasets. First, we introduce LCIdb, a curated, large-sized dataset of DTIs, offering extensive coverage of both the molecule and druggable protein spaces. Notably, LCIdb contains a much higher number of molecules than publicly available benchmarks, expanding coverage of the molecule space. Second, we propose Komet (Kronecker Optimized METhod), a DTI prediction pipeline designed for scalability without compromising performance. Komet leverages a three-step framework, incorporating efficient computation choices tailored for large datasets and involving the Nystrom approximation. Specifically, Komet employs a Kronecker interaction module for (molecule, protein) pairs, which efficiently captures determinants in DTIs, and whose structure allows for reduced computational complexity and quasi-Newton optimization, ensuring that the model can handle large training sets, without compromising on performance. Our method is implemented in open-source software, leveraging GPU parallel computation for efficiency. We demonstrate the interest of our pipeline on various datasets, showing that Komet displays superior scalability and prediction performance compared to state-of-the-art deep learning approaches. Additionally, we illustrate the generalization properties of Komet by showing its performance on an external dataset, and on the publicly available[L][H] benchmark designed for scaffold hopping problems. Komet is available open source at https://komet.readthedocs.io and all datasets, including LCIdb, can be found at https://zenodo.org/records/10731712.

bioinformatics↗

bioRxiv2022

Representation and quantification Of Module Activity from omics data with rROMA

The efficiency of analyzing high-throughput data in systems biology has been demonstrated in numerous studies, where molecular data, such as transcriptomics and proteomics, offers great opportunities for understanding the complexity of biological processes. One important aspect of data analysis in systems biology is the shift from a reductionist approach that focuses on individual components to a more integrative perspective that considers the system as a whole, where the emphasis shifted from differential expression of individual genes to determining the activity of gene sets. Here, we present the rROMA software package for fast and accurate computation of the activity of gene sets with coordinated expression. The rROMA package incorporates significant improvements in the calculation algorithm, along with the implementation of several functions for statistical analysis and visualizing results. These additions greatly expand the packages capabilities and offer valuable tools for data analysis and interpretation. It is an open-source package available on github at: www.github.com/sysbio-curie/rROMA. Based on publicly available transcriptomic datasets, we applied rROMA to cystic fibrosis, highlighting biological mechanisms potentially involved in the establishment and progression of the disease and the associated genes. Results indicate that rROMA can detect disease-related active signaling pathways using transcriptomic and proteomic data. The results notably identified a significant mechanism relevant to cystic fibrosis, raised awareness of a possible bias related to cell culture, and uncovered an intriguing gene that warrants further investigation. Contact: loredana.martignetti@curie.fr

systems biology↗

bioRxiv2019

Evaluation of network architecture and data augmentation methods for deep learning in chemogenomics

AO_SCPLOWBSTRACTC_SCPLOWAmong virtual screening methods that have been developed to facilitate the drug discovery process, chemogenomics presents the particularity to tackle the question of predicting ligands for proteins, at at scales both in the protein and chemical spaces. Therefore, in addition to to predict drug candidates for a given therapeutic protein target, like more classical ligand-based or receptor-based methods do, chemogenomics can also predict off-targets at the proteome level, and therefore, identify potential side-effects or drug repositioning opportunities. In this study, we study and compare machine-learning and deep learning approaches for chemogenomics, that are applicable to screen large sets of compounds against large sets of druggable proteins. State-of-the-art drug chemogenomics methods rely on expert-based chemical and protein descriptors or similarity measures. The recent development of deep learning approaches enabled to design algorithms that learn numerical abstract representations of molecular graphs and protein sequences in an end-to-end fashion, i.e., so that the learnt features optimise the objective function of the drug-target interaction prediction task. In this paper, we address drug-target interaction prediction at the druggable proteome-level, with what we define as the chemogenomic neuron network. This network consists of a feed-forward neuron network taking as input the combination of molecular and protein representations learnt by molecular graph and protein sequence encoders. We first propose a standard formulation of this chemogenomic neuron network. Then, we compare the performances of the standard chemogenomic network to reference deep learning or shallow (machine-learning without deep learning) methods. In particular, we show that such a representation learning approach is competitive with state-of-the-art chemogenomics with shallow methods, but not ultimately superior. We evaluate the most promising neuron network architectures and data augmentation techniques, such as multi-view and transfer learning, to improve the prediction performance of the chemogenomic network. Our results shed new insights on the design of chemogenomics approaches based on representation learning algorithms. Most importantly, we conclude from our observations that a promising research direction is to integrate heterogeneous sources of data such as various bioactivity datasets, or independently, multiple molecule and protein attribute views, instead of focusing on sophisticated, yet intuitively relevant, encoders neuron network architecture.

bioinformatics↗

bioRxiv2018

LOTUS: a Single- and Multitask Machine Learning Algorithm for the Prediction of Cancer Driver Genes

Cancer driver genes, i.e., oncogenes and tumor suppressor genes, are involved in the acquisition of important functions in tumors, providing a selective growth advantage, allowing uncontrolled proliferation and avoiding apoptosis. It is therefore important to identify these driver genes, both for the fundamental understanding of cancer and to help finding new therapeutic targets. Although the most frequently mutated driver genes have been identified, it is believed that many more remain to be discovered, particularly for driver genes specific to some cancer types.\n\nIn this paper we propose a new computational method called LOTUS to predict new driver genes. LOTUS is a machine-learning based approach which allows to integrate various types of data in a versatile manner, including informations about gene mutations and protein-protein interactions. In addition, LOTUS can predict cancer driver genes in a pan-cancer setting as well as for specific cancer types, using a multitask learning strategy to share information across cancer types.\n\nWe empirically show that LOTUS outperforms three other state-of-the-art driver gene prediction methods, both in terms of intrinsic consistency and prediction accuracy, and provide predictions of new cancer genes across many cancer types.\n\nAuthor summaryCancer development is driven by mutations and dysfunction of important, so-called cancer driver genes, that could be targeted by targeted therapies. While a number of such cancer genes have already been identified, it is believed that many more remain to be discovered. To help prioritize experimental investigations of candidate genes, several computational methods have been proposed to rank promising candidates based on their mutations in large cohorts of cancer cases, or on their interactions with known driver genes in biological networks. We propose LOTUS, a new computational approach to identify genes with high oncogenic potential. LOTUS implements a machine learning approach to learn an oncogenic potential score from known driver genes, and brings two novelties compared to existing methods. First, it allows to easily combine heterogeneous informations into the scoring function, which we illustrate by learning a scoring function from both known mutations in large cancer cohorts and interactions in biological networks. Second, using a multitask learning strategy, it can predict different driver genes for different cancer types, while sharing information between them to improve the prediction for every type. We provide experimental results showing that LOTUS significantly outperforms several state-of-the-art cancer gene prediction softwares.

bioinformatics↗

bioRxiv2017

Efficient Multi-task chemogenomics for drug specificity prediction

Adverse drug reactions, also called side effects, range from mild to fatal clinical events and significantly affect the quality of care. Among other causes, side effects occur when drugs bind to proteins other than their intended target. As experimentally testing drug specificity against the entire proteome is out of reach, we investigate the application of chemogenomics approaches. We formulate the study of drug specificity as a problem of predicting interactions between drugs and proteins at the proteome scale. We build several benchmark datasets, and propose NN-MT, a multi-task Support Vector Machine (SVM) algorithm that is trained on a limited number of data points, in order to solve the computational issues or proteome-wide SVM for chemogenomics. We compare NN-MT to different state-of-the-art methods, and show that its prediction performances are similar or better, at an efficient calculation cost. Compared to its competitors, the proposed method is particularly efficient to predict (protein, ligand) interactions in the difficult double-orphan case, i.e. when no interactions are previously known for the protein nor for the ligand. The NN-MT algorithm appears to be a good default method providing state-of-the-art or better performances, in a wide range of prediction scenarii that are considered in the present study: proteome-wide prediction, protein family prediction, test (protein, ligand) pairs dissimilar to pairs in the train set, and orphan cases.

bioinformatics↗

bioRxiv2017

Kernel multitask regression for toxicogenetics

The development of high-throughput in vitro assays to study quantitatively the toxicity of chemical compounds on genetically characterized human-derived cell lines paves the way to predictive toxicogenetics, where one would be able to predict the toxicity of any particular compound on any particular individual. In this paper we present a machine learning-based approach for that purpose, kernel multitask regression (KMR), which combines chemical characterizations of molecular compounds with genetic and transcriptomic characterizations of cell lines to predict the toxicity of a given compound on a given cell line. We demonstrate the relevance of the method on the recent DREAM8 Toxicogenetics challenge, where it ranked among the best state-of-the-art models, and discuss the importance of choosing good descriptors for cell lines and chemicals.

bioinformatics↗