detecting-inter-individual-contamination-mismatches-multiomics-next

Detecting Inter-Individual Contamination and Mismatches in Multiomics Next-Generation Sequencing Data

High-throughput processing of patient biosamples by next-generation sequencing and the comparison of molecular data with patient-level and sample-level ...

Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd

Many bioinformatics methods have been proposed for reducing the complexity of large gene or protein networks into relevant subnetworks or modules. Yet, how such methods compare to each other in terms of their ability to identify disease-relevant modules in different types of network remains poorly understood. We launched the 'Disease Module Identification DREAM Challenge', an open competition to comprehensively assess module identification methods across diverse protein-protein interaction, signaling, gene co-expression, homology and cancer-gene networks. Predicted network modules were tested for association with complex traits and diseases using a unique collection of 180 genome-wide association studies. Our robust assessment of 75 module identification methods reveals top-performing algorithms, which recover complementary trait-associated modules. We find that most of these modules correspond to core disease-relevant pathways, which often comprise therapeutic targets. This community challenge establishes biologically interpretable benchmarks, tools and guidelines for molecular network analysis to study human disease biology.

Assessment of network module identification across complex diseases.

We define a disease module as a partition of a molecular network whose components are jointly associated with one or several diseases or risk factors thereof. Identification of such modules, across different types of networks, has great potential for elucidating disease mechanisms and establishing new powerful biomarkers. To this end, we launched the 'Disease Module Identification (DMI) DREAM Challenge', a community effort to build and evaluate unsupervised molecular network modularization algorithms. Here, we present MONET, a toolbox providing easy and unified access to the three top-performing methods from the DMI DREAM Challenge for the bioinformatics community.

MONET: a toolbox integrating top-performing methods for network modularization.

Melanoma is a highly malignant skin cancer with high propensity to metastasize and develop drug resistance, making it a difficult cancer to treat. Current therapies targeting BRAF (V600) mutations are initially effective, but eventually tumors overcome drug sensitivity and reoccur. This process is accomplished in part by reactivating alternate signaling networks that reinstate melanoma proliferative and survival capacity, mostly through reprogramming of receptor tyrosine kinase (RTK) signaling. Evidence indicates that the discoidin domain receptors (DDRs), a set of RTKs that signal in response to collagen, are part of the kinome network that confer drug resistance. We previously reported that DDR1 is expressed in melanomas, where it can promote tumor malignancy in mouse models of melanoma, and thus, DDR1 could be a promising target to overcome drug resistance. In this review, we summarize the current knowledge on DDRs in melanoma and their implication for therapy, with emphasis in resistance to MAPK inhibitors.

Discoidin Domain Receptors in Melanoma: Potential Therapeutic Targets to Overcome MAPK Inhibitor Resistance.

Next-generation cancer immunotherapies are designed to broaden the therapeutic repertoire by targeting new immune checkpoints including lymphocyte-activation gene 3 (LAG-3) and T cell immunoglobulin and mucin-domain containing-3 (TIM-3). Yet, the molecular and cellular mechanisms by which either receptor functions to mediate its inhibitory effects are still poorly understood. Similarly, little is known on the differential effects of dual, compared with single, checkpoint inhibition.

Deciphering molecular and cellular ex vivo responses to bispecific antibodies PD1-TIM3 and PD1-LAG3 in human tumors.

Understanding protein function and developing molecular therapies require deciphering the cell types in which proteins act as well as the interactions between proteins. However, modeling protein interactions across biological contexts remains challenging for existing algorithms. Here, we introduce Pinnacle, a geometric deep learning approach that generates context-aware protein representations. Leveraging a multi-organ single-cell atlas, Pinnacle learns on contextualized protein interaction networks to produce 394,760 protein representations from 156 cell type contexts across 24 tissues. Pinnacle's embedding space reflects cellular and tissue organization, enabling zero-shot retrieval of the tissue hierarchy. Pretrained protein representations can be adapted for downstream tasks: enhancing 3D structure-based representations for resolving immuno-oncological protein interactions, and investigating drugs' effects across cell types. Pinnacle outperforms state-of-the-art models in nominating therapeutic targets for rheumatoid arthritis and inflammatory bowel diseases, and pinpoints cell type contexts with higher predictive capability than context-free models. Pinnacle's ability to adjust its outputs based on the context in which it operates paves way for large-scale context-specific predictions in biology.

Contextual AI models for single-cell protein biology.

Understanding protein function and developing molecular therapies require deciphering the cell types in which proteins act as well as the interactions between proteins. However, modeling protein interactions across biological contexts remains challenging for existing algorithms. Here we introduce PINNACLE, a geometric deep learning approach that generates context-aware protein representations. Leveraging a multiorgan single-cell atlas, PINNACLE learns on contextualized protein interaction networks to produce 394,760 protein representations from 156 cell type contexts across 24 tissues. PINNACLE's embedding space reflects cellular and tissue organization, enabling zero-shot retrieval of the tissue hierarchy. Pretrained protein representations can be adapted for downstream tasks: enhancing 3D structure-based representations for resolving immuno-oncological protein interactions, and investigating drugs' effects across cell types. PINNACLE outperforms state-of-the-art models in nominating therapeutic targets for rheumatoid arthritis and inflammatory bowel diseases and pinpoints cell type contexts with higher predictive capability than context-free models. PINNACLE's ability to adjust its outputs on the basis of the context in which it operates paves the way for large-scale context-specific predictions in biology.

Bromodomain and extra-terminal domain (BET) inhibitors (BETi) have demonstrated epigenetic modulation capabilities, specifically in transcriptional repression of oncogenic pathways. Preclinical assays suggest that BETi potentially attenuates the PD1/PD-L1 immune checkpoint axis, supporting its combination with immunomodulatory agents.

Immune modulation in solid tumors: a phase 1b study of RO6870810 (BET inhibitor) and atezolizumab (PD-L1 inhibitor).

Hematologic adverse events are common dose-limiting toxicities in drug development. Classical animal models for preclinical safety assessment of immunotherapies are often limited due to insufficient cross-reactivity with non-human homologous proteins, immune system differences, and ethical considerations. Therefore, we evaluate a human bone marrow (BM) microphysiological system (MPS) for its ability to predict expected hematopoietic liabilities of immunotherapeutics. The BM-MPS consists of a closed microfluidic circuit containing a ceramic scaffold covered with human mesenchymal stromal cells and populated with human BM-derived CD34+ cells in chemically defined growth factor-enriched media. The model supports on-chip differentiation of erythroid, myeloid and NK cells from CD34+ cells over 31 days. The hematopoietic lineage balance and output is responsive to pro-inflammatory factors and cytokines. Treatment with a transferrin receptor-targeting IgG1 antibody results in inhibition of on-chip erythropoiesis. The immunocompetence of the chip is established by the addition of peripheral blood T cells in a fully autologous setup. Treatment with T cell bispecific antibodies induces T cell activation and target cell killing consistent with expected on-target off-tumor toxicities. In conclusion, this study provides a proof-of-concept that this BM-MPS is applicable for in vitro hematopoietic safety profiling of immunotherapeutics.

A microfluidic bone marrow chip for the safety profiling of biologics in pre-clinical drug development.

Immune checkpoint inhibitors have become standard care across many cancers, but most patients do not respond. Predicting response remains challenging due to complex tumor-immune interactions and the poor generalizability of current biomarkers and models. Predictors such as tumor mutational burden, PD-L1 expression, and transcriptomic signatures often fail across cancer types, therapies, and clinical settings. There is a clear need for a robust, interpretable model that captures shared immune response principles and adapts to diverse clinical contexts. We present Compass, a foundation model for predicting immunotherapy response from pan-cancer transcriptomic data using a concept bottleneck architecture. Compass encodes tumor gene expression through 44 biologically grounded immune concepts representing immune cell states, tumor-microenvironment interactions, and signaling pathways. Trained on 10,184 tumors across 33 cancer types, Compass outperforms 22 baseline methods in 16 independent clinical cohorts spanning seven cancers and six immune checkpoint inhibitors, increasing precision by 8.5%, Matthews correlation coefficient by 12.3%, and area under the precision-recall curve by 15.7%, with minimal or no additional training. The model generalizes to unseen cancer types and treatments, supporting indication selection and patient stratification in early-phase clinical trials. Survival analysis shows that Compass-stratified responders have significantly longer overall survival (hazard ratio = 4.7,  < 0.0001). Personalized response maps link gene expression to immune concepts, revealing distinct mechanisms of response and resistance. For example, among immune-inflamed non-responders, Compass identifies distinct resistance programs involving TGF-  signaling, endothelial exclusion, CD4+ T cell dysfunction, and B cell deficiency. By combining mechanistic interpretability with transfer learning, Compass provides mechanistic insights into treatment response variability, supports clinical decision-making, and informs trial design.

Generalizable AI predicts immunotherapy outcomes across cancers and treatments.

Monitoring neutrophil gene expression is a powerful tool for understanding disease mechanisms, developing diagnostics, enhancing therapies, and optimizing clinical trials. Neutrophils are sensitive to the processing, storage, and transportation steps that are involved in clinical sample analysis. This study evaluates the capabilities of technologies from 10× Genomics, PARSE Biosciences, and HIVE (Honeycomb Biotechnologies) to generate single-cell RNA sequencing (scRNA-seq) data from human blood-derived neutrophils. Our comparative analysis shows that all methods produced high-quality data, importantly capturing the transcriptomes of neutrophils. Here, we establish a reliable scRNA-seq workflow for neutrophils in clinical trials: we offer guidelines on sample collection to preserve RNA quality and demonstrate how each method performs in capturing sensitive cell populations in clinical practice.

Comparison of single-cell RNA-seq methods to enable transcriptome profiling of neutrophils in clinical samples.

Patients with stage III melanoma are at high risk of relapse. The NADINA trial evaluating neoadjuvant nivolumab plus ipilimumab and the SWOG-1801 trial evaluating neoadjuvant pembrolizumab have demonstrated superior clinical outcomes with neoadjuvant versus adjuvant checkpoint inhibition. Morpheus-Melanoma was a phase 1b/2, randomized umbrella trial evaluating tobemstomig (anti-PD-1/anti-LAG-3 bispecific antibody; n = 40), tobemstomig plus tiragolumab (anti-TIGIT monoclonal antibody; n = 20) and atezolizumab (PD-L1-targeting monoclonal antibody) plus tiragolumab (n = 20) versus nivolumab (anti-PD-1 monoclonal antibody) plus ipilimumab (anti-CTLA-4 monoclonal antibody; n = 22) in stage III melanoma. The primary endpoint was pathological response by independent pathological review. Additional endpoints included safety and exploratory biomarkers. Here tobemstomig showed a similar pathological response rate (pRR) versus nivolumab plus ipilimumab (80.0% (32/40) versus 77.3% (17/22)); major pathological responses were less frequent with tobemstomig versus nivolumab plus ipilimumab treatment (62.5% (25/40) versus 72.7% (16/22)). Tobemstomig plus tiragolumab and atezolizumab plus tiragolumab showed a lower pRR versus nivolumab plus ipilimumab (60.0% (12/20) and 45.0% (9/20) versus 77.3% (17/22), respectively). Tobemstomig demonstrated improved safety versus nivolumab plus ipilimumab, with 2.5% (1/40) and 22.7% (5/22) of patients experiencing grade 3 or higher treatment-related adverse events (TRAEs), respectively, and 0% (0/40) and 13.6% (3/22) of patients discontinuing treatment due to TRAEs, respectively. Grade 3 or higher TRAEs were reported by 15% (3/20) of patients in the tobemstomig plus tiragolumab arm and by no patients in the atezolizumab plus tiragolumab arm. Baseline CD8 and CD3 tumor-infiltrating T cell density, IFNγ pathway and effector T cell gene expression, tumor mutational burden and pre-surgery circulating tumor DNA correlated with pathological response across treatments. In conclusion, in the Morpheus-Melanoma study, tobemstomig demonstrated a similar pathological response and improved safety profile versus nivolumab plus ipilimumab in patients with resectable stage III melanoma. ClinicalTrials.gov identifier: NCT05116202 .

Neoadjuvant PD-1 and LAG-3-targeting bispecific antibody and other immune checkpoint inhibitor combinations in resectable melanoma: the randomized phase 1b/2 Morpheus-Melanoma trial.

Article	Year
Assessment of network module identification across complex diseases. Nature methods\| PubMed ID: 31471613	2019
MONET: a toolbox integrating top-performing methods for network modularization. Bioinformatics (Oxford, England)\| PubMed ID: 32271874	2020
Discoidin Domain Receptors in Melanoma: Potential Therapeutic Targets to Overcome MAPK Inhibitor Resistance. Frontiers in oncology\| PubMed ID: 33014862	2020
Deciphering molecular and cellular ex vivo responses to bispecific antibodies PD1-TIM3 and PD1-LAG3 in human tumors. Journal for immunotherapy of cancer\| PubMed ID: 36319064	2022
Contextual AI models for single-cell protein biology. bioRxiv : the preprint server for biology\| PubMed ID: 37503080	2024
Contextual AI models for single-cell protein biology. Nature methods\| PubMed ID: 39039335	2024
Immune modulation in solid tumors: a phase 1b study of RO6870810 (BET inhibitor) and atezolizumab (PD-L1 inhibitor). BMC cancer\| PubMed ID: 40102759	2025
A microfluidic bone marrow chip for the safety profiling of biologics in pre-clinical drug development. Communications biology\| PubMed ID: 40374778	2025
Generalizable AI predicts immunotherapy outcomes across cancers and treatments. medRxiv : the preprint server for health sciences\| PubMed ID: 40385399	2025
Comparison of single-cell RNA-seq methods to enable transcriptome profiling of neutrophils in clinical samples. Cell reports methods\| PubMed ID: 40957400	2025
Neoadjuvant PD-1 and LAG-3-targeting bispecific antibody and other immune checkpoint inhibitor combinations in resectable melanoma: the randomized phase 1b/2 Morpheus-Melanoma trial. Nature medicine\| PubMed ID: 40993242	2025

Daniel Marbach

Biography

JoVE Journal Publications

Other Publications