Difference between revisions of "Paul"
(→Paul's Notes) |
(→Paul's Notes) |
||
| Line 4: | Line 4: | ||
| − | + | Understanding the Galaxy Data: | |
| − | + | We care about: Fold Change, P-value | |
| − | |||
| − | |||
| − | + | ||
| − | + | Think about total gene expression values and gene functions. Be aware that large gene expression values can be paired with low fold change, but that cn still be significant. | |
| − | + | ||
| + | In the data input, the first data set (numerator) and second data set (denominator) order is important to sign of fold change. | ||
| + | |||
| + | |||
| + | Possible tests: M WT vs P WT, disomic vs trisomic | ||
| + | |||
| + | |||
| + | truncation for excel ensembl gene numbers: =left(A1,18) | ||
| + | |||
| + | Potentially helpful DEseq paper: | ||
| + | http://genomebiology.biomedcentral.com/articles/10.1186/gb-2010-11-10-r106 | ||
| + | https://www.bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.pdf | ||
| + | |||
| + | DEseq2: uses negative binomial distribution to model gene expression | ||
| + | "We assume that the number of reads in sample j that are assigned to gene i can be modeled by a negative binomial (NB) distribution, | ||
| + | Kij~NB(μij,σ2ij)" | ||
Revision as of 15:57, 7 February 2017
Paul's Notes
Understanding the Galaxy Data:
We care about: Fold Change, P-value
Think about total gene expression values and gene functions. Be aware that large gene expression values can be paired with low fold change, but that cn still be significant.
In the data input, the first data set (numerator) and second data set (denominator) order is important to sign of fold change.
Possible tests: M WT vs P WT, disomic vs trisomic
truncation for excel ensembl gene numbers: =left(A1,18)
Potentially helpful DEseq paper: http://genomebiology.biomedcentral.com/articles/10.1186/gb-2010-11-10-r106
https://www.bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.pdf
DEseq2: uses negative binomial distribution to model gene expression "We assume that the number of reads in sample j that are assigned to gene i can be modeled by a negative binomial (NB) distribution, Kij~NB(μij,σ2ij)"
