February 2018

• Spring 2018 Notebook

Week 1 (2/5-2/9)

• Adjusted significance values for edge deletions
• Created a figure to illustrate which edge deletions caused significant changes and which caused no changes

Research Meeting

• Went over abstracts
• Submitted app to the symposium

Week 2 (2/12-2/16)

• began work on poster for the symposium, mostly the background info
• will be using spreadsheet that Maggie made to calculate minMSE (also for use in the future)

Research Meeting

• all abstracts are submitted
• turned in application and dues for Tri Beta

Week 3 (2/19-2/23)

• calculated MSE:minMSE for each deletion network
  • only one was significant
• looked at differences in edge weights in each deletion network compared to the intact network

Research Meeting

• clustering deletions to see which are most similar or most different to the intact network
• work with Brandon to make a heat map of the edge weight changes in each deletion

Week 4 (2/26-3/2)

• worked on figures for poster

Research Meeting

• went over poster figures with Dr. Fitzpatrick
• Notes
  • unweighted db5 network
    • explain more about what it is
  • LSE:minLSE graph
    • sort from smallest to largest
    • color db5 line with a different color
    • fit improves in 5 cases
    • dHMO1-CIN5: what does it look like over time with and without that edge. CIN5 vs. time with and without HMO1-CIN5 edge
    • dHMO1-YOX1: yox1 vs. time with and without MSN2, and with and without HMO1
    • dZAP1-ACE2: ACE2 vs. time with and without the ZAP1 edge
    • Put the mean of the data and standard error on the graphs
    • Also show a deletion vs time where the model got worse
  • MSE:minSME ration and Production Rates
    • may not need the whole picture, maybe just a table to talk about the t tests
  • Some type of MSE:minMSE visual
    • experiment with things, run them by Dr. D and Dr. F
  • Edge Heat Map (table legend to be reformatted)
    • The edge deletions that change a lot are associated with central genes (lots of edges going to and from them: CIN5, MSN2).
    • Perhaps have not gotten the dynamics of GCR2-MSN2 right quite yet
    • Interesting to illustrate where these things are in the network
    • HMO1-CIN5 and ZAP1-ACE2 potentially have a pretty big ripple effect.
    • HMO1-YOX1 and MSN2-YOX1 end at a dead end.