search

kdahlquist / GRNmap

Gene Regulatory Network modeling and parameter estimation
BSD 3-Clause "New" or "Revised" License
4 stars 3 forks source link

Data Analysis Team tasks for Week of 10/24 #266

Closed kdahlquist closed 7 years ago

kdahlquist commented 8 years ago

The tasks for this week:

Note that nobody seems to be working on #170, so I changed the priority to 0.5 on that one.

maggie-oneil commented 8 years ago

Worked with @khorstmann to create a guide for the functions of the SBEToolbox, specifically the view functions, along with random networks "small world" and "ring-lattice." To find powerpoint guide, please see @khorstmann 's comment for the document.

khorstmann commented 8 years ago

Maggie and I ran through virtually every program on SBEToolbox with the 4-node sample network we talked about yesterday. We weren't able to get to the 21-gene network as it took us much longer to get through it than expected. On Monday we can scroll through everything and decide which things to run on our larger test to see how those work. We also need to add on more information on the downloading, installation, etc of the program so that we can maybe use the powerpoint for documentation of future usage of the program.

Powerpoint can't be uploaded here because it is too large, but here's the link to my lab notebook, where it is uploaded under the Oct 25 page http://www.openwetware.org/wiki/Kristen_M._Horstmann_Lab_Notebook

kdahlquist commented 8 years ago

Sorry, I forgot to stop by after lab yesterday because I was in a hurry to get home. I've looked over the PowerPoint--it looks like there will be a lot to talk about next week.

kdahlquist commented 8 years ago

I've booked the first floor LSB conference room (118) for next Monday's (10/31) lab meeting so that we can look at the PowerPoints on the big screen. See you all there at 3:00.

khorstmann commented 8 years ago

http://www.openwetware.org/wiki/Kristen_M._Horstmann_Lab_Notebook#October_25.2C_2016 @bengfitzpatrick here is the link to my Lab Notebook. Under the entry for Oct 25, there is the powerpoint we showed today

khorstmann commented 8 years ago

Searched through https://github.com/kdahlquist/DahlquistLab/tree/master/data/GRNmap_input_workbooks where we were directed to find the networks to test SBE Toolbox on, but couldn't find the weighted networks. So far had just been working on the unweighted networks when we realized. Could @bklein7 and @Nwilli31 please let us know where their weighted networks are? Thanks!

khorstmann commented 8 years ago

@maggie-oneil an through all the stats and the adjacency for the 3 unweighted networks. POwerpoint can be found in the Nov 1 entry here: http://www.openwetware.org/wiki/Kristen_M._Horstmann_Lab_Notebook#November_1.2C_2016

Determined that the stats (or, at least shortest path, and therefore, betweenness centrality) assume the networks are undirected, which is a bit of an issue for our networks. I'm sure we will discuss in the meeting on monday how to continue

Nwilli31 commented 8 years ago

I looked at the wt output sheet parameters again and realized that I recorded the incorrect numbers for the Penalty term and the LSE. I had them switched, so I changed the numbers on the powerpoint. I reuploaded the file on my openwetware page and on the Dahlquist Repository.

kdahlquist commented 8 years ago

The Data Analysis team meetings will take place in LSB 118 for the rest of the semester.

bklein7 commented 7 years ago

All 5 output sheets from the round of model runs done for issue #265 have been uploaded to the Dahlquist Lab Repository, so @khorstmann and @maggie-oneil can access them there.

maggie-oneil commented 7 years ago

@khorstmann and I input two weighted networks to see if the SBEToolbox took the weights into account when running statistics of the graphs. After running both it appears as though the toolbox does not take into account the weights or direction of edges. We then looked into both yEd and gephi to see if those graphing tools would be able to take direction and weight into account, but found both to be lacking in a cursory review of the programs. The powerpoint for this week is in PDF format and linked directly to this comment. Weighted SBEToolbox.pdf

kdahlquist commented 7 years ago

It looks like there are solutions out there for finding the shortest path in a weighted, directed network, but it might be harder to find "off the shelf" software.

A Google search, turned up this: http://www.geeksforgeeks.org/shortest-path-exactly-k-edges-directed-weighted-graph/

maggie-oneil commented 7 years ago

@khorstmann and I looked into the graph applications and stats applications of the program Gephi and it appears to be more useful and user-friendly than SBEToolbox. We used both a simple, unweighted network and the weighted output sheet for GLN3 in our analysis of the program. The powerpoint of our findings is attached directly to this comment. Giphe.pptx

kdahlquist commented 7 years ago

Just wanted to confirm that the data analysis team is still meeting on Monday 11/21 at 3:00 in the LSB conference room. See you there!

kdahlquist commented 7 years ago

Closing this and starting a new task list.