The approach can also discover novel connections between diseases, which may have an impact on clinical therapy possibilities.
Many brain illnesses have complex hereditary and environmental risk factors, making classification challenging. Furthermore, the symptoms of many brain illnesses overlap. Parkinson’s disease and dementia with Lewy bodies, for example, are both neurodegenerative illnesses characterised by muscle tremors and rigidity, as well as certain cognitive and behavioural signs. Misdiagnosis is widespread in this and other comparable circumstances, and it can have major ramifications for patient treatment. Another strategy is to categorise brain illnesses based on gene activity. The current study looked at disease transcriptomes, which are collections of RNA transcripts from afflicted brain areas, for 40 distinct brain diseases.
The researchers found that this system could classify brain diseases into five primary groups based on where disease-risk genes were active in the brain and in which cell types. In addition to confirming known relationships among diseases, disease transcriptome analysis was able to find previously unknown relationships among diseases. For example, language development disorders, obsessive-compulsive disorder, and temporal lobe epilepsy were all classified into group 3, meaning that despite their very different symptoms, their corresponding genes are active in the same brain regions and in the same cell types. Brain diseases classified as neurodegenerative, movement-related, and psychiatric are the most difficult to diagnose because of the overlap in symptoms that change over time. A transcriptome is thus an additional tool that could be used for more accurate early diagnoses.
Zeighami adds, “Analysis of the transcription patterns of risk genes for human brain disease reveals characteristic expression signatures across brain anatomy. These can be used to compare and aggregate diseases, providing associations that often differ from conventional phenotypic classification.”