and log-likelihood scores. While it uses data other than expression
data to find associations, a lot of the data it does use are in fact
expression data sets and as such we are including AraNet in the
coexpression tool category.
1. Go to https://www.inetbio.org/aranet. Click on “Networksearch.” On the following page, click “Query Option 1. Find
new members of a pathway.”
2. Under “Gene Set,” enter “AT3G24650.” You can leave
Organism on “Auto-Detection.” Click “Submit.”
3. You will need to enable Flash on the next page to see the
interactive views (that are powered by Cytoscape Web—in
Chrome click on the “Secure” beside the URL to “Allow”).
The first view becomes more interesting with multiple genes.
Scroll down to the one element table entry on “AT3G24650.”
You will note that the associated GO terms for ABI3 seem to
make sense.
4. Keep scrolling to the second interactive view (try viewing in a
new window if you cannot see anything). You can see how
many genes AraNet associates with ABI3 (SIS10) by counting
how many entries are in the table.
5. A nice feature of AraNet is use it to look for associations in
common between genes of interest. Start a new “Networksearch” ! “Find new members of a pathway.” Under “Gene
Set,” this time enter both ABI3 “At3g24650” and LEC1
“At1g21970.” You can leave Organism on “Auto-Detection.”
Click “Submit.”
6. Once again, scroll down to the second network view. Take a
look at the node labeled “FG.” You will see that it is associated
with both ABI3 and LEC1. In this way, we can use AraNet to
help us understand connections between genes or gene
products.
7. We can also use the “Query option II. Infer functions from
network neighbors” under “Network-search” to try to infer
functions based on the GO annotations of the genes associated
with our gene or genes of interest—check it out!
3.5.4 AtCAST2
Instead of looking for genes with similar expression patterns across
a set of samples, AtCAST allows you to ask questions about correlations between experiments, that is, it performs the correlation
calculation by generating a vector across all genes for a given
experiment and then asking “is there another sample where those
genes also show similar signatures of expression.” It does this for all
data sets in its database, but importantly allows you to also ask for
your own data set, perhaps generated from a mutant-of-unknownfunction line or a chemical-of-unknown-mode-of-action
Arabidopsis Bioinformatics
45
data to find associations, a lot of the data it does use are in fact
expression data sets and as such we are including AraNet in the
coexpression tool category.
1. Go to https://www.inetbio.org/aranet. Click on “Networksearch.” On the following page, click “Query Option 1. Find
new members of a pathway.”
2. Under “Gene Set,” enter “AT3G24650.” You can leave
Organism on “Auto-Detection.” Click “Submit.”
3. You will need to enable Flash on the next page to see the
interactive views (that are powered by Cytoscape Web—in
Chrome click on the “Secure” beside the URL to “Allow”).
The first view becomes more interesting with multiple genes.
Scroll down to the one element table entry on “AT3G24650.”
You will note that the associated GO terms for ABI3 seem to
make sense.
4. Keep scrolling to the second interactive view (try viewing in a
new window if you cannot see anything). You can see how
many genes AraNet associates with ABI3 (SIS10) by counting
how many entries are in the table.
5. A nice feature of AraNet is use it to look for associations in
common between genes of interest. Start a new “Networksearch” ! “Find new members of a pathway.” Under “Gene
Set,” this time enter both ABI3 “At3g24650” and LEC1
“At1g21970.” You can leave Organism on “Auto-Detection.”
Click “Submit.”
6. Once again, scroll down to the second network view. Take a
look at the node labeled “FG.” You will see that it is associated
with both ABI3 and LEC1. In this way, we can use AraNet to
help us understand connections between genes or gene
products.
7. We can also use the “Query option II. Infer functions from
network neighbors” under “Network-search” to try to infer
functions based on the GO annotations of the genes associated
with our gene or genes of interest—check it out!
3.5.4 AtCAST2
Instead of looking for genes with similar expression patterns across
a set of samples, AtCAST allows you to ask questions about correlations between experiments, that is, it performs the correlation
calculation by generating a vector across all genes for a given
experiment and then asking “is there another sample where those
genes also show similar signatures of expression.” It does this for all
data sets in its database, but importantly allows you to also ask for
your own data set, perhaps generated from a mutant-of-unknownfunction line or a chemical-of-unknown-mode-of-action
Arabidopsis Bioinformatics
45