Almudena Espín-Pérez

Emerging high-throughput sequencing technologies promise lower costs and higher scalability, yet their performance on archival clinical samples remains poorly characterized. Here, we benchmarked Ultima Genomics UG100 against Illumina Novaseq platforms across single-nuclei RNA-seq (snRNA-seq), whole-transcriptome (WTS), whole-exome (WES), and whole-genome sequencing (WGS) using FFPE tissues from oncologic and immune-mediated diseases. Across matched samples, we systematically assessed data quality, coverage profiles, error spectra, variant concordance and transcriptomic reproducibility. UG100 produced highly comparable results to Illumina, capturing key oncogenic and immune-related transcripts, accurately resolving cellular composition in snRNA-seq, and maintaining sensitivity for lowly expressed genes, despite characteristic insertion-biased indels and modest differences in multi-mapping reads. Discrepancies were subtle, largely limited to pseudogene and non-coding transcripts, and did not affect pathway-level conclusions. Ultima UG100 platform prioritized high precision and reduced low-frequency artifacts, offering a cleaner but more conservative variant-calling profile compared to the more sensitive, yet noisier, Illumina/DRAGEN workflow. This multimodal, clinically oriented assessment provides the first comprehensive evaluation of UG100, demonstrating its translational utility in population-scale genomics, and highlighting the potential for emerging sequencing technologies to lower the cost of biomedical research and clinical diagnostics.

Accurate prediction of patient outcomes remains a major challenge in oncology. While recent machine learning (ML) approaches often rely on bulk omics lacking spatial resolution or histology-based multiple instance learning (MIL), spatial transcriptomics (SpT) provides a unique opportunity to capture both molecular content and tissue architecture. However, no generalizable ML framework has yet been established to exploit SpT for patient-level outcomes. We present SpaMIL, a flexible and interpretable MIL framework designed for SpT, with a distillation strategy that enables deployment for hematoxylin and eosin (H&E) slides alone. We evaluate the framework by predicting survival from glioblastoma (GBM) patients, a clinically compelling setting given its aggressiveness with a median survival of only 15 months and the lack of prognostic clinical variables. We analyzed 76 GBM cases from the MOSAIC dataset: 43 with matched SpT, H&E, single-nucleus RNA-seq (scRNA-seq), bulk RNA-seq, and clinical variables, and 33 with H&E for external validation. We developed two main architectures: abMIL, tailored to SpT’s spatial molecular structure, and MabMIL, which distills SpT-derived representations into H&E. Model interpretability was achieved through a Shapley-based framework linking prognostic predictions to cell-type compositions via SpT deconvolution. In benchmarking across the five GBM MOSAIC modalities, SpT-based abMIL achieved unprecedented prognostic accuracy (median C-index: 0.72, standard deviation: 0.04), outperforming all other modalities, including established clinical predictors. PCA and deconvolution-based SpT representations surpassed recent foundation models, suggesting the need for further research on SpT foundation models. Our interpretability analysis highlighted malignant and non-malignant cell subpopulations associated with favorable or poor prognosis, consistent with recent reports. Finally, MabMIL maintained strong performance while enabling H&E-only deployment, with improved condorance index over H&E-only baselines in both internal (0.59 vs. 0.57) and external (0.62 vs. 0.55) cohorts.

Background: Glioblastoma (GBM), the most aggressive primary brain tumor, has a median survival of approximately 15 months. Twenty percent of patients survive beyond three years, but known clinical factors like age, performance status, resection extent, and MGMT promoter methylation status do not fully explain the observed outcomes. Objective: Our objective was to identify novel histology derived biomarkers associated with end-of-spectrum overall survival (OS) to provide novel biological insight with a translational potential. Methods: We analyzed a total of 748 GBM patients from 3 different cohorts, uniquely enriched in long survivors (n=98 with overall survival (OS) > 5y including n=196 with OS≥3y), with clinical data and H&E slides obtained from the primary tumor at baseline. We propose an interpretable machine learning (ML) methodology for the discovery of histological biomarkers. Our method learned to segment each H&E slide into three distinct regions associated with long-term survival, short-term survival, and non-informative tissue. We characterized these regions by integrating unsupervised learning, nuclei segmentation, blood vessels detection, pathologist annotations, and multimodal data including spatial transcriptomics from n=31 patients of the GBM MOSAIC dataset to discover fully interpretable biomarkers. Results: Our OS prediction model using histology and clinical data as input achieved an area under the curve (AUC) of 0.85 for the classification of patients between OSPLIN2 expression and lipid accumulation- is consistently enriched in histological areas predictive of poor prognosis. Conclusion: Our interpretable ML methodology identified a novel prognostic impact of biological processes and cell types according to distinct tumor regions of GBM. These results pave the way for spatially-informed biomarkers to improve risk stratification and for personalized spatially-targeted therapeutic strategies. Key highlights: Our ML model identified histological biomarkers predicting prognosis independently from known clinical factors The region of lowly infiltrated white matter enriched in malignant cells including a mesenchymal-like phenotype is predictive of poor prognosis Angiogenesis is increased in areas predictive of long survival in main non-necrotic tumor regions. The subtype of macrophages expressing PLIN2 and associated with increased lipid metabolism was associated with poor prognosis in all GBM regions. Highlights:

Bulk RNA sequencing provides an averaged gene expression profile of the numerous cells in a tissue sample, obscuring critical information about cellular heterogeneity. Computational deconvolution methods can estimate cell type proportions in bulk samples, but current approaches can lack precision in key scenarios due to simplistic statistical assumptions, limited modeling of cell-type heterogeneity and poor handling of rare populations. We present MixupVI, a deep generative model that learns representations of single-cell transcriptomic data and introduces a mixup-based regularization to enable reference-free deconvolution of bulk samples. Our method creates a latent representation with an additive property, where the representation of a pseudobulk sample corresponds to the weighted sum of its constituent cell types. We demonstrate how MixupVI enables accurate estimation of cell type proportions through benchmarking on pseudobulks simulated from a large immune single-cell atlas. To support reproducibility and foster progress in the field, we also release PyDeconv, a Python library that implements multiple state-of-the-art deconvolution algorithms and provides a comprehensive benchmark on simulated pseudobulk datasets.

PCBs are classified as xenoestrogens and carcinogens and their health risks may be sex-specific. To identify potential sex-specific responses to PCB-exposure we established gene expression profiles in a population study subdivided into females and males. Gene expression profiles were determined in a study population consisting of 512 subjects from the EnviroGenomarkers project, 217 subjects who developed lymphoma and 295 controls were selected in later life. We ran linear mixed models in order to find associations between gene expression and exposure to PCBs, while correcting for confounders, in particular distribution of white blood cells (WBC), as well as random effects. The analysis was subdivided according to sex and development of lymphoma in later life. The changes in gene expression as a result of exposure to the six studied PCB congeners were sex- and WBC type specific. The relatively large number of genes that are significantly associated with PCB-exposure in the female subpopulation already indicates different biological response mechanisms to PCBs between the two sexes. The interaction analysis between different PCBs and WBCs provides only a small overlap between sexes. In males, cancer-related pathways and in females immune system-related pathways are identified in association with PCBs and WBCs. Future lymphoma cases and controls for both sexes show different responses to the interaction of PCBs with WBCs, suggesting a role of the immune system in PCB-related cancer development.

Batch effects are technical sources of variation introduced by the necessity of conducting gene expression analyses on different dates due to the large number of biological samples in population-based studies. The aim of this study is to evaluate the performances of linear mixed models (LMM) and Combat in batch effect removal. We also assessed the utility of adding quality control samples in the study design as technical replicates. In order to do so, we simulated gene expression data by adding “treatment” and batch effects to a real gene expression dataset. The performances of LMM and Combat, with and without quality control samples, are assessed in terms of sensitivity and specificity while correcting for the batch effect using a wide range of effect sizes, statistical noise, sample sizes and level of balanced/unbalanced designs. The simulations showed small differences among LMM and Combat. LMM identifies stronger relationships between big effect sizes and gene expression than Combat, while Combat identifies in general more true and false positives than LMM. However, these small differences can still be relevant depending on the research goal. When any of these methods are applied, quality control samples did not reduce the batch effect, showing no added value for including them in the study design.

Environmental Health Perspectives is an Open Access journal published by the National Institute of Environmental Health Sciences. Environmental Health Perspectives is an Open Access journal published by the National Institute of Environmental Health Sciences.

Considering genetic variability in population studies focusing on the health risk assessment of exposure to environmental carcinogens may provide improved insights in individual environmental cancer risks. Therefore, the current study aims to determine the impact of genetic polymorphisms on the relationship between exposure and gene expression, by identifying exposure-dependently coregulated genes and genetic pathways. Statistical analysis based on mixed models, was performed to relate gene expression data from 134 subjects to exposure measurements of multiple carcinogens, 28 polymorphisms, age, sex and biomarkers of cancer risk. We evaluated the combined exposure to cadmium, lead, polychlorinated biphenyls, p,p′-dichlorodiphenyldichloroethylene, hexachlorobenzene and 1-OH-pyrene, and the outcome was biologically interpreted by using ConsensusPathDB, thereby focusing on carcinogenesis-related pathways. We found generic and carcinogenesis-related pathways deregulated in both sexes, but males showed a stronger transcriptome response than females. We highlighted NOTCH1, CBR1, ITGB3, ITGA4, ADI1, HES1, NCOA2 and SMARCA2 in view of their direct link with cancer development. Two of these, NOTCH1 and ITGB3, are also known to respond to PCBs and cadmium chloride exposure in rodents and to lead in humans. Subjects carrying a high number of risk alleles appear more responsive to combined carcinogen exposure with respect to the induced expression of some of these cancer-related genes, which may be indicative of increased cancer risk as a consequence of environmental factors.