Difference between revisions of "February 4, 2016"
From GcatWiki
(→Correlation Activity) |
|||
| (One intermediate revision by the same user not shown) | |||
| Line 1: | Line 1: | ||
== Classwork == | == Classwork == | ||
Consider how signaling cascades often involve a G-protein, receptor, and ligand. The Burmese Pythons do not always feed, so it is likely that the transcripts for these genes are not always present. Therefore, we are looking for a '''''transcription factor activator''''' because these genes likely turn on a multitude of other genes. | Consider how signaling cascades often involve a G-protein, receptor, and ligand. The Burmese Pythons do not always feed, so it is likely that the transcripts for these genes are not always present. Therefore, we are looking for a '''''transcription factor activator''''' because these genes likely turn on a multitude of other genes. | ||
| + | |||
'''''Moving forward, we need to:''''' | '''''Moving forward, we need to:''''' | ||
| Line 18: | Line 19: | ||
*Slope indicates if data is positively or negatively related, or if there is no correlation. | *Slope indicates if data is positively or negatively related, or if there is no correlation. | ||
*We can practice correlating gene expression across gene samples. | *We can practice correlating gene expression across gene samples. | ||
| + | *Correlation deals with rate of change, NOT magnitude of change. | ||
| + | *Subtle changes occurring within a single sample can dramatically change the correlation coefficient. | ||
| + | **Biologically, it is not uncommon to have a gene that does not change much. | ||
| + | |||
| + | '''''We need to understand "who" the outlier is in our research.''''' | ||
| + | |||
| − | + | === Questions to Consider: === | |
| + | *Knowing subtle changes within a sample can dramatically change the correlation coefficient, should we include all three snakes from each category in a cluster? | ||
| + | *How do we cluster the most effectively? | ||
Latest revision as of 17:42, 13 February 2016
Classwork
Consider how signaling cascades often involve a G-protein, receptor, and ligand. The Burmese Pythons do not always feed, so it is likely that the transcripts for these genes are not always present. Therefore, we are looking for a transcription factor activator because these genes likely turn on a multitude of other genes.
Moving forward, we need to:
- Validate our 12 samples and compare them to one another.
- Identify a housekeeping gene in the Small Intestine mucosa.
- Determine a grouping on the dedrogram and set a threshold.
- Identify genes that will distinguish fed-state from fasted-state.
Questions to Consider:
- When working with large data sets, how do we separate out what is statistically and biologically interesting?
Correlation Activity
- R2 value indicates how well the trend line explains the correlation of data. R2 is the square of the correlation coefficient.
- Slope indicates if data is positively or negatively related, or if there is no correlation.
- We can practice correlating gene expression across gene samples.
- Correlation deals with rate of change, NOT magnitude of change.
- Subtle changes occurring within a single sample can dramatically change the correlation coefficient.
- Biologically, it is not uncommon to have a gene that does not change much.
We need to understand "who" the outlier is in our research.
Questions to Consider:
- Knowing subtle changes within a sample can dramatically change the correlation coefficient, should we include all three snakes from each category in a cluster?
- How do we cluster the most effectively?
