Biology subjects

Barkmann, F.

Publications and source records attributed to Barkmann, F..

7 recordsLinked to original sources

bioRxiv2025

Sparse Autoencoders Reveal Interpretable Features in Single-Cell Foundation Models

Single-cell foundation models (scFMs) hold promise for applications in cell type annotation, data integration, and prediction of the effects of cell perturbations, but their internal mechanisms remain poorly understood. We investigate the structure of these models by training sparse autoencoders (SAEs) on the hidden representations of three widely used scFMs: scGPT, scFoundation, and Geneformer.The learned features reveal diverse and complex biological and technical signals, which emerge even in pre-trained models. We also observe that the encoding of this information differs between scFMs with distinct training protocols and architectures. Finally, we demonstrate that SAE-derived features are functionally related to model behavior and can be intervened upon. Suppressing batch-associated features reduces unwanted technical variation and improves data integration while preserving the core biological signal. Activating drug-encoding features steers control cells toward drug-perturbed states in a concentration-dependent manner. These findings provide a path toward more interpretable and controllable single-cell foundation models.

bioinformatics↗

bioRxiv2024

Benchmarking Self-Supervised Learning for Single-Cell Data

Self-supervised learning (SSL) has emerged as a powerful approach for learning biologically meaningful representations of single-cell data. To establish best practices in this domain, we present a comprehensive benchmark evaluating eight SSL methods across three downstream tasks and eight datasets, with various data augmentation strategies. Our results demonstrate that SimCLR and VICReg consistently outperform other methods across different tasks. Furthermore, we identify random masking as the most effective augmentation technique. This benchmark provides valuable insights into the application of SSL to single-cell data analysis, bridging the gap between SSL and single-cell biology.

bioinformatics↗

bioRxiv2024

CancerFoundation: A single-cell RNA sequencing foundation model to decipher drug resistance in cancer

We present CancerFoundation, a novel single-cell RNA-seq foundation model (scFM) trained exclusively on malignant cells. Despite being trained on only one million total cells, a fraction of the data used by existing models, CancerFoundation outperforms other scFMs in key tasks such as zero-shot batch integration and drug response prediction. During training, we employ tissue and technologyaware oversampling and domain-invariant training to enhance performance on underrepresented cancer types and sequencing technologies. We propose survival prediction as a new downstream task to evaluate the generalizability of single-cell foundation models to bulk RNA data and their applicability to patient stratification. CancerFoundation demonstrates superior batch integration performance and shows significant improvements in predicting drug responses for both unseen cell lines and drugs. These results highlight the potential of focused, smaller foundation models in advancing drug discovery and our understanding of cancer biology. Our code is available here1.

bioinformatics↗

bioRxiv2023

ANS: Adjusted Neighborhood Scoring to improve assessment of gene signatures in single-cell RNA-seq data

In the field of single-cell RNA sequencing (scRNA-seq), gene signature scoring is integral for pinpointing and characterizing distinct cell populations. However, challenges arise in ensuring the robustness and comparability of scores across various gene signatures and across different batches and conditions. Here, we evaluated the stability of established methods such as Scanpy, UCell, and JASMINE in the context of scoring cells of different types and states. On eight cancer and healthy scRNA-seq datasets, we reported that none of the existing methods provide fair gene signature scores that can be used in unsupervised cell state annotation based on the highest signature values. Addressing this challenge, we introduced a new scoring method, the Adjusted Neighbourhood Scoring (ANS), that builds on the traditional Scanpy method and improves the handling of the control gene sets. We further exemplified the usability of ANS scoring in differentiating between cancer-associated fibroblasts and malignant cells undergoing epithelial-mesenchymal transition (EMT) in four cancer types, and evidenced excellent classification performance (AUCPR train: 0.95-0.99, AUCPR test: 0.91-0.99). In summary, our research introduces ANS as a robust and deterministic scoring approach that enables the comparison of diverse gene signatures and score-based annotation of cell types and states. The results of our study contribute to the development of more accurate and reliable methods for analyzing scRNA-seq data.

bioinformatics↗

bioRxiv2022

CanSig: Discovering de novo shared transcriptional programs in single cancer cells

Single-cell RNA sequencing (scRNA-seq) facilitates the discovery of gene signatures that define cell states across patients, which could be used in patient stratification and drug discovery. However, the lack of standardization in computational methodologies to analyse these data impedes the reproducibility of signature detection. To address this, we developed CanSig, a comprehensive benchmarking tool that evaluates methods for identifying transcriptional signatures in cancer. CanSig integrates metrics for batch correction and biological signal conservation with a gene signature correlation metric to score according to rediscovery, cross-dataset reproducibility, and clinical relevance. We applied CanSig to ten methods and to ten scRNA-seq datasets from four human cancer types--glioblastoma, breast cancer, lung adenocarcinoma, and cutaneous squamous cell carcinoma-- representing 116 patients and 105,000 malignant cells. Our results identify BBKNN as a leading method. We showed that the signatures identified with these methods correlate with clinically relevant outcomes, including patient survival and lymph node metastasis. Thus, CanSig establishes a standardized framework for reproducible cancer transcriptomics analysis.

bioinformatics↗

bioRxiv2022

scROSHI - robust supervised hierarchical identification of single cells

Identifying cell types based on expression profiles is a pillar of single cell analysis. Existing machine-learning methods identify predictive features from annotated training data, which are often not available in early-stage studies. This can lead to overfitting and inferior performance when applied to new data. To address these challenges we present scROSHI, which utilizes previously obtained cell type-specific gene sets and does not require training or the existence of annotated data. By respecting the hierarchical nature of cell type relationships and assigning cells consecutively to more specialized identities, excellent prediction performance is achieved. In a benchmark based on publicly available PBMC data sets, scROSHI outperforms competing methods when training data are limited or the diversity between experiments is large.

bioinformatics↗

bioRxiv2019

The transcription factor Zt107320 affects the dimorphic switch, growth and virulence of the fungal wheat pathogen Zymoseptoria tritici

Zymoseptoria tritici is a filamentous fungus causing Septoria tritici blotch in wheat. The pathogen has a narrow host range and infections of grasses other than susceptible wheat are blocked early after stomatal penetration. During these abortive infections the fungus shows a markedly different expression pattern. However, the underlying mechanisms causing differential gene expression during host and non-host interaction are largely unknown, but likely include transcriptional regulators responsible for the onset of an infection program in compatible hosts. In the rice blast pathogen Magnaporthe oryzae, MoCOD1, a member of the fungal Zn(II)2Cys6 transcription factor family, has been shown to directly affect pathogenicity. Here, we analyse the role of the putative transcription factor Zt107320, a homolog of MoCOD1, during infection of compatible and incompatible hosts by Z. tritici. We show for the first time that Zt107320 is differentially expressed in host versus non-host infections and that lower expression corresponds to an incompatible infection of non-hosts. Applying reverse genetics approaches we further show that Zt107320 regulates the dimorphic switch as well as the growth rate of Z. tritici and affects fungal cell wall composition in vitro. Moreover, {Delta}Zt107320 mutants showed reduced virulence during compatible infections of wheat. We conclude that Zt107320 directly influences pathogen fitness and propose that Zt107320 regulates growth processes and pathogenicity during infection. Our results suggest that this putative transcription factor is involved in discriminating compatible and non-compatible infections.

microbiology↗