 |
 |
GEO help: Mouse over screen elements for information. |
|
| Status |
Public on Oct 27, 2011 |
| Title |
Temporal Dynamics and Genetic Control of Transcription in the Human Prefrontal Cortex |
| Organism |
Homo sapiens |
| Experiment type |
Expression profiling by array
|
| Summary |
In order to gain a global molecular perspective on how the human genome gives rise to the brain, we explore the temporal dynamics and genetic control of transcription in human dorsolateral prefrontal cortex in an extensive series of brain tissue from fetal development through aging. We discover a wave of gene expression changes occurring during fetal development which are reversed in early postnatal life. One half century later in life, this phenomenon is mirrored in aging as well as neurodegeneration. While we identify thousands of robust associations of individual genetic polymorphisms with gene expression, we also demonstrate that there is no association between the total extent of genetic differences between subjects and the global similarity of their transcriptional profiles. Hence, the human genome produces a consistent molecular architecture in the prefrontal cortex, despite millions of genetic differences across individuals and races. To enable further discovery, this entire dataset is freely available (GEO).
The gene expression from this study can be interrogated gene-by-gene with a biologist-friendly web application available at http://braincloud.jhmi.edu/plots/.
|
| |
|
| Overall design |
RNA from 269 human prefrontal cortex samples ranging from fetal development (negative ages) through aging (80 years) were analyzed on custom 2-color microarrays from the National Human Genome Research Institute (NHGRI) microarray core facility using a reference RNA comprised of a pool of all samples. Subjects had no severe neuropathology nor neurological or neuropsychiatric diagnoses.
Preprocessed/normalized data for exploration is contained in the Series Matrix file (see "Download Family" section below). These data are very flexible in that they are well suited to diverse downstream analyses, provided that a well-designed "cleaning" method is first applied. Care should be taken in using a "cleaning" method that is particularly suited to the exact downstream analysis that will be performed. As there is significant biological and technical/experimental heterogeneity in the data (albeit less here with this 2-color technology than is usually associated with large 1-color studies), we recommend using surrogate variable analysis (SVA; PMID:22257669) or other data "cleaning" methodology such as ComBat (PMID:16632515) to focus the data on a particular question.
Another version of the data tuned particularly to identifying canonical patterns of gene expression across the lifespan (at the expense of individual variation) can be downloaded from the "Supplementary file" section below ("GSE30272_ExprsMtxCleanedN269_31SVN.txt"). Here we expand on the previous modeling (Colantuoni 2011, PMID: 22031444) of age patterns across the lifespan in several key ways (manuscript in preparation): 1) We applied splines to capture non-linear gene expression effects while ensuring patterns of gene expression are continuous across the lifespan. The previous analysis used age by decade interaction terms, which are not necessarily continuous. 2) We estimated and adjusted for a much higher number of SVs. The previous analysis used only 2 SVs, here we allowed SVA to automatically determine this number: 31 SVs were used. This much increased "cleaning" further tuned this dataset to age effects. Hence, this newly processed data should only be used for the estimation of canonical, mean patterns of expression across the lifetime. 3) We regressed out SVs while allowing the effects of age and mean gene expression (the intercept) to remain in the data. Previously, SVs were regressed out while ignoring possible correlation between SVs and age, potentially obscuring some age effects. Specifically, using SVA, we employ a 2nd degree basis spline with knots at birth, 1, 10, 20, and 50 years [8 degrees of freedom], i.e. a curve fit to expression across age within each age range between these knots. Each model also allowed an offset at birth, because there were no samples in the third trimester of fetal life. The exact model matrix used to implement this spline analysis in SVA has been included as a supplementary file for download ("GSE30272_ModelMatrix.txt.gz" in the "Supplementary file" section below) and may be useful for incorporating this model into further statistical analysis when beginning from the "uncleaned" data. The individual values for all 31 SVs used in our analysis are also included as a supplementary file ("GSE30272_SVN31Matrix.txt.gz" in the "Supplementary file" section below).
For easy, platform-independent access to images of individual gene expression profiles across human DLPFC development derived from this data processing/cleaning (i.e. data from the Supplementary file "GSE30272_ExprsMtxCleanedN269_31SVN.txt"), visit: http://braincloud.jhmi.edu/plots/.
Raw red-green intensity data are contained in the supplementary file "GSE30272_RAW.tar" (see "Supplementary file" section below). These would need to be converted to log ratios, normalized, and "cleaned" prior to any biological analysis.
While we do not recommend using them for data exploration, the data that are processed as originally described in Colantuoni 2011 (PMID: 22031444) can be re-created by downloading the preprocessed, but not "cleaned" data in the Series Matrix file (see "Download Family" section below) and adjusting for (in regression models) or regressing out (for visualization) the 2 surrogate variables contained in the array/sample annotation in this same Series Matrix file.
SNP genotype data for these subjects can be found in dbGaP under Study ID *phs000417.v1.p1* http://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000417.v1.p1.
|
| Web link |
http://braincloud.jhmi.edu/plots/
|
| |
|
| Contributor(s) |
Colantuoni C, Lipska BK, Ye T, Hyde TM, Tao R, Leek JT, Colantuoni EA, Elkahloun AG, Herman MM, Weinberger DR, Jaffe A, Kleinman JE |
| Citation(s) |
22031444 |
| |
| Submission date |
Jun 28, 2011 |
| Last update date |
Apr 26, 2018 |
| Contact name |
Carlo Colantuoni |
| E-mail(s) |
ccolantu@jhmi.edu
|
| Phone |
4104931439
|
| Organization name |
Johns Hopkins Univ. School of Medicine
|
| Department |
Neurology
|
| Street address |
733 N Broadway
|
| City |
Baltimore |
| State/province |
MD |
| ZIP/Postal code |
21205 |
| Country |
USA |
| |
|
| Platforms (1) |
| GPL4611 |
Illumina Human 49K Oligo array (HEEBO-7 set) |
|
| Samples (269)
|
|
| Relations |
| BioProject |
PRJNA143735 |
| Supplementary file |
Size |
Download |
File type/resource |
| GSE30272_ExprsMtxCleanedN269_31SVN.txt.gz |
67.0 Mb |
(ftp)(http) |
TXT |
| GSE30272_ModelMatrix.txt.gz |
7.6 Kb |
(ftp)(http) |
TXT |
| GSE30272_RAW.tar |
1.0 Gb |
(http)(custom) |
TAR (of TXT) |
| GSE30272_SVN31Matrix.txt.gz |
72.9 Kb |
(ftp)(http) |
TXT |
| Processed data included within Sample table |
|
|
|
|
 |
Less...
More...