Difference between revisions of "Introduction to Microarray analysis"
From Organic Design wiki
m (→Data issues: ffs -moron) |
m |
||
| Line 11: | Line 11: | ||
====Data issues==== | ====Data issues==== | ||
*In the past statistics was developed for n >>p | *In the past statistics was developed for n >>p | ||
| − | :<font color= | + | :<font color="red">n observations, p variables</font> |
*Gene expression data n<<p | *Gene expression data n<<p | ||
| − | **Thousands of measured genes (p) | + | **<font color="red">Thousands of measured genes (p)</font> |
| − | **Small number of biological replicate slides (n) | + | **<font color="red">Small number of biological replicate slides (n)</font> |
*Gene expression data can be highly correlated | *Gene expression data can be highly correlated | ||
| − | **groups of genes are regulated in the same way | + | **<font color="red">groups of genes are regulated in the same way</font> |
*Data not normally distributed | *Data not normally distributed | ||
| − | **log transform highly skewed intensity data | + | **<font color="red">log transform highly skewed intensity data</font> |
---- | ---- | ||
Revision as of 01:55, 14 March 2006
Overview of experimental process
- competitive hybridization to spotted oligo/cDNA transcripts
- Interested in genes that change between treatments
- → differential expression versus equivalent expression
Statistical analysis process
Data issues
- In the past statistics was developed for n >>p
- n observations, p variables
- Gene expression data n<<p
- Thousands of measured genes (p)
- Small number of biological replicate slides (n)
- Gene expression data can be highly correlated
- groups of genes are regulated in the same way
- Data not normally distributed
- log transform highly skewed intensity data
Scratch pad
- A flow diagram for analysis
- Recap of cDNA microarrays (slide 3)
- Microarray data issues (slide 4)
- Microarray data issues (continued)
- Large amount of data (GPR/JPEG file size)
- Subjective
- Need a log of what was done so someone else can quickly reroduce the results
- → Reproducible research (someone else can understand/reproduce the results) (McGintys talk)
- Analysis process
- R resources/contributed guides (including downloading)
- R tutorial of basics (objects/indexing/functions)
