Characterizing the developing human brain transcriptome at single-base resolution

点击翻译按钮获取中文摘要

 DESCRIPTION (provided by applicant): Human post-mortem brain tissue provides a valuable resource for identifying expression-based determinants of development and subsequent dysregulation in brain disorders (Kleinman 2011). RNA sequencing (RNA-seq) generates potentially unbiased characterization of the transcriptome, and has now been performed on over 1,000 non-psychiatric human brain samples across the lifespan within our lab (LIBD) and via the BrainSpan project (www.brainspan.org). We propose to combine and re-process these data together to make them more comparable, starting from raw sequencing reads. Our goal is further interrogate the clinical relevance of developmentally dynamic regions of gene expression across brain development. Our preliminary RNAseq data demonstrates extensive developmental regulation of previously unannotated intra- and inter-genic sequence conserved across multiple brain regions in both humans and mice. Base-level analysis of these combined RNA-seq data can greatly improve existing gene annotation databases like Ensembl and UCSC, which currently lack many fetal brain-specific transcripts that we have identified and characterized in our samples. We will perform base-level resolution analyses on both the entire dataset, from first trimester of fetal life through the aged (>85 years), and then secondary analyses exploring differential expression across different brain regions, to identify dynamic expressed sequence. Identified differentially expressed regions (DERs) will be interrogated for clinical significance through enrichment analysis with regard to predefined clinical gene sets, for examples significant loci from genome-wide association studies (GWAS) for brain disorders like schizophrenia, and also by directly associating genetic risk with expression levels within identified DERs. We will further provide the genomic tools to allow researchers with interests in other genes and loci to design their own validation experiments in cell lines and/or primary brain tissue for these clinically relevant loci. A more comprehensive characterization of the human brain transcriptome, leveraging "big data" across two complementary datasets, will be valuable for scientists studying a wide range of developmental processes and brain disorders.