Uncovering relationships between stress and psychiatric traits via multi-omics individual-specific networks (ISNs)

Objectives and research questions

Brief biological introduction

Input data

Methods

Results and aditional information

Authors

Objectives

The primary aim of this study is to investigate the genetic variants that influence stress-induced environmental patterns and to assess the relationship between these environmental and genetic biomarkers with the risk of psychiatric disorders. Specifically, the project seeks to address the following two research questions:

1. How are individuals clustered based on changes observed before and after stimulation with dexamethasone, a synthetic glucocorticoid (a type of steroid) that mimics cortisol, a hormone naturally released during stress?

2. What are the key factors driving these changes?

Brief biological introduction

Dexamethasone, a synthetic glucocorticoid, is commonly used in research to investigate the hypothalamic-pituitary-adrenal (HPA) axis, which regulates the body’s stress response. The dexamethasone suppression test (DST) is a well-established method for assessing HPA axis function. By administering dexamethasone and monitoring the body's ability to suppress cortisol production, researchers can gain insights into stress-related physiological responses.

In psychiatric and stress-related research, the DST is particularly relevant for studying how individuals respond to stress and whether disruptions in cortisol regulation contribute to the development of psychiatric disorders, such as depression and anxiety.

In this study, dexamethasone stimulation is used to:

1. Observe how genetic variants affect the body's stress response.
2. Analyze the environmental and genetic factors influencing individual differences in cortisol regulation, which may be associated with psychiatric risk.

Input data

Participants comprised 196 Caucasian individuals of the Max Planck Institute of Psychiatry (MPIP) in Munich. Of the participants,

• 131 men and 65 women
• 84 (50 men, 34 women) were treated for major dipressive disorders (MDD)
• 112 (81 men, 31 women) were healthy controls with no history of a depressive disorder.

Baseline whole blood samples were obtained at 6 pm after two hours of fasting and abstention from coffee and physical activity. Subjects then received 1.5 mg oral dexamethasone, and a second blood draw was performed at 9 pm, three hours after dexamethasone ingestion.

The available multiomics data for the currect project:

• Methylation
• Gene-expression (for later analysis)
• Genotype
• Phenotype

The following figure displays a flow of the data collection:

Data are stored on the MPIP computational cluster.

Methods

Individual-Specific Network (ISN) Construction

To address the research questions, Individual-Specific Networks (ISNs), which focus on personalized data to model each individual’s unique associations over time, were utilized. ISNs are constructed using multiple measurements for the same individual, allowing capturing temporal dynamics and environmental influences that may differ from the population-level network.

Why ISNs?

1. Personalized Insight: ISNs enable the translation of population-level interpretations to the individual. By focusing on individual-specific edges (connections between nodes representing biomarkers), one can extrapolate from general models to individual dynamics, providing a more granular understanding of how each person’s biological and environmental factors interact.

2. Temporal Focus: ISNs allow to observe each individual's specific associations and changes over time, providing a longitudinal view of how stress-induced patterns develop and evolve within an individual.

How ISNs are Constructed

The ISN methodology involves constructing a network where nodes represent biomarkers, and edges represent individual-specific associations between these biomarkers. In this case, the focus is on individual-specific edges (the node information is available or not), which may differ from the average population network.

The ISNs are generated by:

• Collecting multiple measurements.
• Using these measurements to compute associations between biomarkers for each individual, which are then represented as edges in the network.
• Comparing the network structure pre- and post-stimulation to observe how an individual's stress response modulates these associations.

The graphical representation below outlines the general workflow for constructing ISNs:

The subsequent figure highlights the project-specific ISN construction process, demonstrating how biomarker data is translated into network models that capture individual-specific dynamics:

The ISN approach is key to identifying individual drivers of change and linking genetic variants to stress responses and potential psychiatric disorder risks.

Results and additional information

Results are stored on the MPIP computational cluster.

For additional information on methodology, results and limitations, please refer to the slides.

For further questions and/or suggestions, please contact the owner of this repository.

Authors

Anastasiia Hryhorzhevska