Neural Network-Assisted Humanization of COVID-19 Hamster scRNAseq Data Reveals Matching Severity States in Human Disease

Vincent D Friedrich,Peter Pennitz,Emanuel Wyler,Julia M Adler,Dylan Postmus,Luiz Gustavo Teixeira Alves,Julia Prigann,Fabian Pott,Daria Vladimirova,Thomas Hoefler,Cengiz Goekeri,Markus Landthaler,Christine Goffinet,Antoine-Emmanuel Saliba,Markus Scholz,Martin Witzenrath,Jakob Trimpert,Holger Kirsten,Geraldine Nouailles

DOI: https://doi.org/10.1101/2024.01.11.574849

2024-01-12

Abstract:Translating findings from animal models to human disease is essential for dissecting disease mechanisms, developing and testing precise therapeutic strategies. The coronavirus disease 2019 (COVID-19) pandemic has highlighted this need, particularly for models showing disease severity-dependent immune responses. Single-cell transcriptomics (scRNAseq) is well poised to reveal similarities and differences between species at the molecular and cellular level with unprecedented resolution. However, computational methods enabling detailed matching are still scarce. Here, we provide a structured scRNAseq-based approach that we applied to scRNAseq from blood leukocytes originating from humans and hamsters affected with moderate or severe COVID-19. Integration of COVID-19 patient data with two hamster models that develop moderate (Syrian hamster, ) or severe (Roborovski hamster, ) disease revealed that most cellular states are shared across species. A neural network-based analysis using variational autoencoders quantified the overall transcriptomic similarity across species and severity levels, showing highest similarity between neutrophils of Roborovski hamsters and severe COVID-19 patients, while Syrian hamsters better matched patients with moderate disease, particularly in classical monocytes. We further used transcriptome-wide differential expression analysis to identify which disease stages and cell types display strongest transcriptional changes. Consistently, hamster’s response to COVID-19 was most similar to humans in monocytes and neutrophils. Disease-linked pathways found in all species specifically related to interferon response or inhibition of viral replication. Analysis of candidate genes and signatures supported the results. Our structured neural network-supported workflow could be applied to other diseases, allowing better identification of suitable animal models with similar pathomechanisms across species.

Immunology

What problem does this paper attempt to address?

The aim of this paper is to address the problem of how to match disease states across different species using neural network technology, particularly comparing the severity of COVID-19 disease between humans and hamsters at the single-cell level. Specifically:

1. **Cross-species comparison**: Using single-cell transcriptomics data (scRNA-seq) to study the similarity of disease states between humans and two types of hamsters (Syrian hamsters and Roborovski hamsters).
2. **Disease severity matching**: Quantifying the transcriptomic similarity between different species and disease severities using the Variational Autoencoder (VAE) method, and finding the best-matched disease states.
3. **Validation of method generality**: Ensuring that the proposed neural network method is not only applicable to the training set samples but can also be generalized to unseen datasets.
4. **Differential gene expression analysis**: Identifying which genes and pathways drive similar disease processes in human patients and hamster models.

Through these means, researchers hope to better understand the relationship between animal models and human disease states, thereby providing guidance for future disease research and treatment strategies.