SMLG (Statistical Machine Learning Group) Discussion Forum

Hi Samantha.

Could you include these ER stress genes in the last network? Thank you for your help
Atf3
Derl3
Sdf2l1
Mcm10
Trib3
IRE1
ATF6
PERK
ATF4
XBP1

jliuzzi wrote:Hi Samantha.

Could you include these ER stress genes in the last network? Thank you for your help
Atf3
Derl3
Sdf2l1
Mcm10
Trib3
IRE1
ATF6
PERK
ATF4
XBP1

Here is the first draft of the requested network! Note that 4 of the genes are disconnected from the network because they do not have any neighbors in common with the second degree Markov blanket.
A note on how the networks are generated:

Network generation:
To generate the network, I create a node list by
1. Finding the nodes in the markov blanket of "pizMutation"
2. Finding the nodes in the markov blanket of all nodes in the first degree MB,
3. Manually adding any requested genes not inherently present in the second degree MB to the node list.

I then use the node list to generate a subgraph from the original network.

Let me know if you have any questions/suggestions!

samanthagonzales wrote:

jliuzzi wrote:Hi Samantha.

Could you include these ER stress genes in the last network? Thank you for your help
Atf3
Derl3
Sdf2l1
Mcm10
Trib3
IRE1
ATF6
PERK
ATF4
XBP1

Here is the first draft of the requested network! Note that 4 of the genes are disconnected from the network because they do not have any neighbors in common with the second degree Markov blanket.
A note on how the networks are generated:

Network generation:
To generate the network, I create a node list by
1. Finding the nodes in the markov blanket of "pizMutation"
2. Finding the nodes in the markov blanket of all nodes in the first degree MB,
3. Manually adding any requested genes not inherently present in the second degree MB to the node list.

I then use the node list to generate a subgraph from the original network.

Let me know if you have any questions/suggestions!

Here is an updated draft by adding 8 genes ("Btnl9", "Ccl6", "Cyp2a22", "Ddit4", "Ehhadh", "Gstp3", "Mup3", "Ndrg1") to the network to ensure all nodes are connected:
These genes are overlapping neighbors between the previously disconnected nodes (Hspa8, Xbp1, Derl3, Atg7) and the Markov blanket nodes which were also listed in the annotation reference colorcodefunction.xlsx

I updated the previously posted network by removing 2 of the "connector" genes that weren't actually connecting the previously disconnected nodes to the network. So now there are 6 additional genes that are present only to connect those previously disconnected nodes (Atg7, Hspa8, Xbp1, and Derl3) to the network.

I also generated a "summarized" version of the network by removing all of the red connector genes, but retaining the connection between the disconnected node and the network.
For example, Derl3 connects to Map1lc3b via Btnl9 (Derl3 -> Btnl9 -> Map1lc3b) : In the summarized version, Btnl9 is removed, and an arc pointing from Derl3 to Map1lc3bis added (Derl3 -> Map1lc3b).

Please let me know if you have further suggestions! Thanks