Daniel:Notebook/RNAFISH/2014-12-5
Cell Line Analysis[edit]
I've been using three cell lines to test the RNA FISH, U-87MG, T98-G, and BE2-C. None of the results from the 11/22/2014 test were conclusive, so I also got some data from Dr. Zhang on NCBI's website. This data is meant to see whether or not we have gene expression of the 4 genes I have already (RBFOX3/NeuN, FOXP2, CUX2, and GAD1) in the 4 cell lines, since we just assumed before. Links to the raw data can be found on the cell line pages (links above).
Gene Expression Analysis[edit]
I found two separate types of gene expression data sets. Two data sets (T98-G and U87-MG) were analyzed using Affymetrix microarrays, and one cell line (BE2-C) was analyzed using RNA-Seq.
Affymetrix Microarray Data (T98-G and U87-MG)[edit]
The first challenge of this is to understand the Affymetrix labeling system. With help from DAVID I changed the Affymetrix notation to Entrez Genes and identified which labels (usually several) corresponded to my genes of interest. The data is summarized below.
Gene | Entrez Gene ID | Gene List | U87-MG Values | T98-G Values |
RBFOX3 | 146713 | 229406_at | 1.654966 | 1.756282 |
RBFOX3 | 146713 | 235468_at | 2.087149 | 2.173624 |
RBFOX3 | 146713 | 240975_x_at | 3.09884 | 3.058095 |
FOXP2 | 93986 | 1562139_a_at | 2.320036 | 2.509934 |
FOXP2 | 93986 | 1555516_at | 1.644757 | 1.88334 |
FOXP2 | 93986 | 243278_at | 1.420676 | 2.523345 |
FOXP2 | 93986 | 1552902_a_at | 1.915386 | 2.145663 |
FOXP2 | 93986 | 235201_at | 2.150443 | 3.007517 |
FOXP2 | 93986 | 1555647_a_at | 2.238783 | 2.606263 |
FOXP2 | 93986 | 1555648_at | 1.967491 | 2.01729 |
FOXP2 | 93986 | 1564876_s_at | 3.461954 | 3.740007 |
FOXP2 | 93986 | 1555352_at | 1.765198 | 2.079463 |
CUX2 | 23316 | 1566528_at | 2.388545 | 2.225919 |
CUX2 | 23316 | 213920_at | 1.672316 | 1.586615 |
CUX2 | 23316 | 1566527_at | 3.108661 | 2.981597 |
GAD1 | 2571 | 205278_at | 4.696471 | 1.507815 |
GAD1 | 2571 | 206669_at | 4.935924 | 3.5153 |
GAD1 | 2571 | 206670_s_at | 4.498899 | 2.625633 |
Since this is microarray data, the numbers don't mean a lot on their own (they range from about 1 to 14). So I plotted, for each cell line, the total distribution (boxplot) and also the individual values for each of the four genes.
- U87MG.jpg
- T98G CellLineAnalysis.png
So the U87-MG line might have an enrichment of GAD1, since that is the only gene for either line that appears to be significantly above the mean.
RNASeq Data (BE2-C)[edit]
This data is a little more tame. Basically I just sorted through a large text file until I found the chromosomal locations of each gene (using Gene Cards), then found the transcript name and FPKM. The table of results is below.
Gene | Gene ID | FPKM | Orientation | Confirmed |
RBFOX3 | ENST00000338834.4 | 0 | - | No |
FOXP2 | ENST00000393500.3 | 0.0004195009 | + | No |
CUX2 | ENST00000261726.5 | 0.6117341499 | + | Maybe |
GAD1 | ENST00000358196.3 | 0 | + | No |
This one shows that the BE2-C line may contain appreciable amounts of CUX2. I'm basing this off the gene card for CUX2, which has the following information:
Using the same calculation for this (100*FPKM)^1/2, this would yield a score of 7.8, which for CUX2 seems to be about as high as it goes.