Biology subjects

Schwikowski, B.

Publications and source records attributed to Schwikowski, B..

6 recordsLinked to original sources

bioRxiv2026

Single-Cell Transcriptomic Signatures Enable Stratified Combination Therapy for Platinum-Resistant Ovarian Cancer

In high-grade serous carcinoma (HGSC), extensive intra-tumoral heterogeneity hinders complete eradication and remains a major obstacle to developing combination therapies capable of eliminating subpopulations resistant to standard-of-care treatment. Using single-cell RNA sequencing of 72 samples from 54 HGSC patients spanning treatment-naive, post-neoadjuvant chemotherapy and relapse stages, we established a carboplatin-anchored framework that identifies transcriptional signatures of intrinsic (pre-existing) and adaptive (therapy-induced) resistance in individual tumors and prioritizes mechanistically matched drugs to potentiate carboplatin efficacy. Candidate compounds were ranked by integrating orthogonal resources--viability (GDSC, PRISM) and perturbational transcriptomics (L1000, Perturb-seq)--to reduce context bias. Among 64 candidates, three carboplatin adjuvants enhanced long-term efficacy in patient-derived organoids (PDOs), and pevonedistat further significantly reduced tumor burden in orthotopic xenografts. This tiered validation pipeline--from short-term and long-term PDOs and in vivo orthoptic xenografts--establishes a translational framework linking single cell resistance programs to actionable, tumor-specific, carboplatin-anchored combinations for HGSC.

cancer biology↗

bioRxiv2025

Identification of Monotonically Classifying Pairs of Genes for Ordinal Disease Outcomes

In this study, we extend an existing classification method for identifying pairs of genes whose joint expression is associated with binary outcomes to ordinal multi-class outcomes, such as overall survival or disease progression. Our approach is motivated by the need for interpretable classifiers that can provide insights into the underlying biological mechanisms. It can be easily adapted to different research questions, such as identifying gene pair signatures or functional enrichment. We demonstrate that our method is comparable to state-of-the-art classification approaches in terms of performance, while offering the benefit of higher interpretability and adaptability to solve different research questions. Our evaluation on two real-world use cases in glioblastoma and high-grade serous ovarian carcinoma shows that our approach can effectively predict ordinal outcomes and provide novel biological insights. The code is available at https://github.com/oceanefrqt/MBMC.

systems biology↗

bioRxiv2024

Consensus Clustering for Robust Bioinformatics Analysis

Clustering plays an important role in a multitude of bioinformatics applications, including protein function prediction, population genetics, and gene expression analysis. The results of most clustering algorithms are sensitive to variations of the input data, the clustering algorithm and its parameters, and individual datasets. Consensus clustering (CC) is an extension to clustering algorithms that aims to construct a robust result from those clustering features that are invariant under the above sources of variation. As part of CC, stability scores can provide an idea of the degree of reliability of the resulting clustering. This review structures the CC approaches in the literature into three principal types, introduces and illustrates the concept of stability scores, and illustrates the use of CC in applications to simulated and real-world gene expression datasets. Open-source R implementations for each of these CC algorithms are available in the GitHub repository: https://github.com/behnam-yousefi/ConsensusClustering

bioinformatics↗

bioRxiv2023

Fast Identification of Optimal Monotonic Classifiers

MotivationBivariate monotonic classifiers (BMCs) are based on pairs of input features. Like many other models used for machine learning, they can capture non-linear patterns in high-dimensional data. At the same time, they are simple and easy to interpret. Until now, the use of BMCs on a genome scale was hampered by the high computational complexity of the search for pairs of features with a high leave-one-out performance estimate. ResultsWe introduce the fastBMC algorithm, which drastically speeds up the identification of BMCs. The algorithm is based on a mathematical bound for the BMC performance estimate while maintaining optimality. We show empirically that fastBMC speeds up the computation by a factor of at least 15 already for a small number of features, compared to the traditional approach. For two of the three clinical datasets that we consider here, the resulting possibility of considering much larger sets of features translates into significantly improved classification performance. As an example for the high degree of interpretability of BMCs, we discuss a straightforward interpretation of a BMC glioblastoma survival predictor, an immediate novel biomedical hypothesis, options for biomedical validation, and treatment implications. ConclusionsfastBMC enables the rapid construction of robust and interpretable ensemble models using BMC, facilitating the discovery of interesting gene pairs and their contributions to the underlying biology. AvailabilityWe provide the first open-source implementation for learning BMCs, and an implementation of fastBMC in particular, all in Python, at https://github.com/oceanefrqt/fastBMC.

bioinformatics↗

bioRxiv2023

Capturing the dynamics of microbiomes using individual-specific networks

BackgroundLongitudinal analysis of multivariate individual-specific microbiome profiles over time or across conditions remains a daunting task. The vast majority of statistical tools and methods available to study the microbiota are based upon cross-sectional data. Over the past few years, several attempts have been made to model the dynamics of bacterial species over time or across conditions. However, the field needs novel views on how to incorporate individual-specific microbial associations in temporal analyses when the focus lies on microbial interactions. ResultsHere, we propose a novel data analysis framework, called MNDA, to uncover taxon neighbourhood dynamics that combines representation learning and individual-specific microbiome co-occurrence networks. We show that tracking local neighbourhood dynamics in microbiome interaction or co-occurrence networks can yield complementary information to standard approaches that only use microbial abundances or pairwise microbial interactions. We use cohort data on infants for whom microbiome data was available at 6 and 9 months after birth, as well as information on mode of delivery and diet changes over time. In particular, MNDA-based prediction models outperform traditional prediction models based on individual-specific abundances, and enable the detection of microbes whose neighbourhood dynamics are informative of clinical variables. We further show that similarity analyses of individuals based on microbial neighbourhood dynamics can be used to find subpopulations of individuals with potential relevance to clinical practice. The annotated source code for the MNDA framework can be downloaded from: https://github.com/H2020TranSYS/microbiome_dynamics ConclusionsMNDA extracts information from matched microbiome profiles and opens new avenues to personalized prediction or stratified medicine with temporal microbiome data.

microbiology↗

bioRxiv2021

Differential levels of IFNα subtypes in autoimmunity and viral infection

Type I interferons are essential for host response to viral infections, while dysregulation of their response can result in autoinflammation or autoimmunity. Among IFN (alpha) responses, 13 subtypes exist that signal through the same receptor, but have been reported to have different effector functions. However, the lack of available tools for discriminating these closely related subtypes, in particular at the protein level, has restricted the study of their differential roles in disease. We developed a digital ELISA with specificity and high sensitivity for the IFN2 subtype. Application of this assay, in parallel with our previously described pan-IFN assay, allowed us to study different IFN protein responses following cellular stimulation and in diverse patient cohorts. We observed different ratios of IFN protein responses between viral infection and autoimmune patients. This analysis also revealed a small percentage of autoimmune patients with high IFN2 protein measurements but low pan-IFN measurements. Correlation with an ISG score and functional activity showed that in this small sub group of patients, IFN2 protein measurements did not reflect its biological activity. This unusual phenotype was partly explained by the presence of anti-IFN auto-antibodies in a subset of autoimmune patients. This study reports ultrasensitive assays for the study of IFN proteins in patient samples and highlights the insights that can be obtained from the use of multiple phenotypic readouts in translational and clinical studies.

immunology↗