Transcriptomic subtyping of imaging phenotypes in autism spectrum disorder

• Applicant: Professorin Christine Ecker, Ph.D.
• Subject Area: Biological Psychiatry ; Human Cognitive and Systems Neuroscience
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 551579379

Autism Spectrum Disorder (ASD) is a highly heterogeneous neurodevelopmental condition with a complex etiology and neurobiology. Heterogeneity exists both between individuals who fulfil diagnostic criteria, as well as within individuals across development. Hence, clinical trials, which have largely been unsuccessful so far, typically rely on biologically diverse groups of ASD individuals, operationally defined according to Diagnostic and Statistical Manual of Mental Disorders (DSM-5) and the International Statistical Classification of Diseases and Related Health Problems (ICD-10) diagnostic criteria. The identification of more homogenous biological subgroups is therefore essential for the development of novel treatments based on the molecular mechanisms underpinning ASD, and for the development of personalized (i.e., individually tailored) treatment and intervention strategies. Recent advances in brain imaging transcriptomics now make the identification of new treatment targets and the stratification of individuals according to putative molecular markers possible. Here, publicly available gene expression data (e.g., the Allen Human Brain Atlas) is brought into alignment with Magnetic Resonance Imaging (MRI) data of the brain, to identify genes with a spatial pattern of expression resembling a neuroimaging signature, and so to link macroscopic pathology with microscopic mechanisms. Within the proposed project, we now aim utilize this technique to (i) map imaging phenotypes in ASD to molecular targets, and (ii) to stratify ASD individuals with respect to putative candidate mechanisms. We will leverage existing data from the EU-AIMS Longitudinal European Autism Project (LEAP, N>700) and a local DFG-funded longitudinal sample (N>200) of ASD individuals and TD controls with both neuroimaging and genetics data. The work program will be subdivided into three stages. In Stage 1, we will develop a computational framework for linking complex neuroanatomical imaging phenotypes to genes and genetic pathways. In Stage 2, this framework will be applied to the transcriptomic profiling and subtyping of neuroanatomical ASD imaging phenotypes (LEAP sample). In Stage 3, we aim to validate putative ASD subgroups in an independent sample of individuals (DFG sample), and against clinical and genetic data. The outcomes of the project will feed into several ongoing research initiatives aimed at obtaining regulatory biomarker approval for ASD (e.g., www.eu-aims.eu), and will inform the development of personalized treatment strategies to reduce the high disease burden and unmet clinical needs associated with the condition.

• DFG Programme: Research Grants
• International Connection: France, United Kingdom
• Cooperation Partners: Professor Dr. Thomas Bourgeron; Professor Declan Murphy
• Co-Investigators: Professor Andreas G. Chiocchetti, Ph.D.; Professorin Dr. Christine Margarete Freitag