Biology subjects

Selega, A.

Publications and source records attributed to Selega, A..

6 recordsLinked to original sources

bioRxiv2025

Learning copy number dependent variation in single tumour cell transcriptomes with deep generative models

Large-scale copy number aberrations are a hallmark of many cancers, showing widespread associations with tumour cell phenotypes, therapy resistance, and patient outcomes. Several existing methods can infer copy number profiles from single-cell RNA-sequencing data. However, there is currently no methodology to separate the effects of the copy number state on single tumour cell phenotypes from the copy number-independent variation. We present ISOMERIC, an unsupervised deep learning framework to learn disentangled representations of single-cell expression data that are dependent on or independent of copy number profiles. We apply ISOMERIC to multiple cancer types and examine how copy number shapes transcriptional states. We find distinct associations between copy number-dependent and independent variation and clinical subtypes in multiple cancer types and link such variation to tumour microenvironment phenotypes. Together, this establishes a principled framework for untangling the effects of genetic and intrinsic variation on tumour transcriptomes and the consequences on clinically meaningful phenotypes.

bioinformatics↗

bioRxiv2025

Multi-view deep learning of highly multiplexed imaging data improves association of cell states with clinical outcomes

Highly multiplexed imaging technologies can quantify the expression of 10s-100s of proteins and their modifications at sub-cellular resolution. Applications to tissue specimens across a broad array of pathologies have enabled patient subtyping in terms of novel cell states and spatial cell communities and linked their prevalence and overall architecture to patient outcomes. While highly multiplexed imaging analysis workflows typically summarize each cell in terms of its post-segmentation mean expression, additional cellular information can be quantified including cell morphology, sub-cellular expression patterns, and spatial cellular context, ultimately giving a multi-modal view of each cell. While deep latent variable models such as variational autoencoders (VAEs) are well established for other multi-modal single-cell assays, their ability to integrate these multiple views of a cell from highly multiplexed imaging data remains largely unknown. Here, we explore the abilities of multi-modal VAEs to learn unified latent cellular representations from multiple views of each single-cell quantified from highly multiplexed imaging, including mean expression, morphology, sub-cellular protein co-localization, and spatial cellular context, while conditioning on technical and batch specific effects. We first discuss strategies for training and hyperparameter optimization of such models across a set of highly multiplexed imaging datasets of breast and melanoma cancers. We next quantify the relevance of the integrated multi-modal latent space in predicting patient-specific clinical outcomes and demonstrate competitive performance across a range of outcomes compared to existing baselines. Then, we explore the ability of the cellular representations to learn cellular phenotypes that align with known cell types and find that the expression-specific latent representation can identify clusters more closely aligned with known lymphocyte and epithelial subpopulations compared to established workflows. Finally, we test the ability of each representation to perform consistent batch integration across different cohorts and discuss the considerations on taking known background variables into account.

bioinformatics↗

bioRxiv2025

Multi-modal Disentanglement of Spatial Transcriptomics and Histopathology Imaging

Spatially-resolved expression profiling data has revolutionized biological research with multiple emerging clinical applications. Spatial transcriptomic assays are often jointly measured with histopathology imaging data, which is frequently used for diagnosing and staging various diseases. However, determining the extent to which the spatial transcriptomic and histopathology data represent overlapping or unique sources of variation is challenging, particularly given the myriad of factors influencing both, including expression variation, spatial context, tissue morphology, and batch effects. Here, we view this challenge as multi-modal disentanglement and develop an evaluation framework. We introduce SpatialDIVA, a disentanglement technique for jointly measured spatially resolved transcriptomics and histopathology data. We demonstrate that SpatialDIVA outperforms baseline techniques in disentangling salient factors of variation in curated pathologist-annotated multi-sample colorectal and pancreatic cancer cohorts. Further, SpatialDIVA removes batch effects from multi-modal data, allows for factor covariance analysis, and yields actionable biological insights through a novel conditional multi-modal generation method. The SpatialDIVA model, evaluation code, and datasets are available at https://github.com/hsmaan/SpatialDIVA.

genomics↗

bioRxiv2024

Segmentation error aware clustering for highly multiplexed imaging

Spatial protein expression technologies can map cellular content and organization by simultaneously quantifying the expression of >40 proteins at subcellular resolution within intact tissue sections and cell lines. However, necessary image segmentation to single cells is challenging and error prone, easily confounding the interpretation of cellular phenotypes and cell clusters. To address these limitations, we present STARLING, a novel probabilistic machine learning model designed to quantify cell populations from spatial protein expression data while accounting for segmentation errors. To evaluate performance we developed a comprehensive benchmarking workflow by generating highly multiplexed imaging data of cell line pellet standards with controlled cell content and marker expression and additionally established a novel score to quantify the biological plausibility of discovered cellular phenotypes on patient derived tissue sections. Moreover, we generate spatial expression data of the human tonsil - a densely packed tissue prone to segmentation errors - and demonstrate cellular states captured by STARLING identify known cell types not visible with other methods and enable quantification of intra- and inter- individual heterogeneity. STARLING is available at https://github.com/camlab-bioml/starling.

bioinformatics↗

bioRxiv2024

Beyond benchmarking: towards predictive models of dataset-specific single-cell RNA-seq pipeline performance

The advent of single-cell RNA-sequencing (scRNA-seq) has driven significant computational methods development for all steps in the scRNA-seq data analysis pipeline, including filtering, normalization, and clustering. The large number of methods and their resulting parameter combinations has created a combinatorial set of possible pipelines to analyze scRNA-seq data, which leads to the obvious question: which is best? Several benchmarking studies have sought to compare methods to answer this, but frequently find variable performance depending on dataset and pipeline characteristics. Alternatively, the large number of publicly available scRNA-seq datasets along with advances in supervised machine learning raise a tantalizing possibility: could the optimal pipeline be predicted for a given dataset? Here we begin to answer this question by applying 288 scRNA-seq analysis pipelines to 86 datasets and quantifying pipeline success via a range of measures evaluating cluster purity and biological plausibility. We build supervised machine learning models to predict pipeline success given a range of dataset and pipeline characteristics. We find both that prediction performance is significantly better than random and that in many cases pipelines predicted to perform well provide clustering outputs similar to expert-annotated cell type labels. Finally, we identify characteristics of scRNA-seq datasets that correlate with strong prediction performance that could guide when such prediction models may be useful.

bioinformatics↗

bioRxiv2022

Multi-objective Bayesian Optimization with Heuristic Objectives for Biomedical and Molecular Data Analysis Workflows

Many practical applications require optimization of multiple, computationally expensive, and possibly competing objectives that are well-suited for multi-objective Bayesian optimization (MOBO) procedures. However, for many types of biomedical data, measures of data analysis workflow success are often heuristic and therefore it is not known a priori which objectives are useful. Thus, MOBO methods that return the full Pareto front may be suboptimal in these cases. Here we propose a novel MOBO method that adaptively updates the scalarization function using properties of the posterior of a multi-output Gaussian process surrogate function. This approach selects useful objectives based on a flexible set of desirable criteria, allowing the functional form of each objective to guide optimization. We demonstrate the qualitative behaviour of our method on toy data and perform proof-of-concept analyses of single-cell RNA sequencing and highly multiplexed imaging datasets.

bioinformatics↗