Details
-
Type:
Task
-
Status: Closed (View Workflow)
-
Priority:
Major
-
Resolution: Done
-
Affects Version/s: None
-
Fix Version/s: None
-
Labels:None
-
Story Points:3
-
Epic Link:
-
Sprint:Spring 5, Spring 6
Description
GOAL:
To use the NEXTFLOW and deseq2 results to get differential gene expression patterns.
- Between 25C and 37C within a variety. How many genes? A list of them. How much over lap btw the 4 plant varieties (upsetR)?
- All genes at 37C each variety.
- Heatmap like Anthony made in their paper.
Original request is below:
Right now, we have the following data:
Heinz 25C expression patterns & numbers in each path
Heinz 37C expression patterns & numbers in each path
Malintka 25C expression patterns & numbers in each path
Malintka 37C expression patterns & numbers in each path
Nagcarlang 25C expression patterns & numbers in each path
Nagcarlang 37C expression patterns & numbers in each path
Tamaulipas 25C expression patterns & numbers in each path
Tamaulipas 37C expression patterns & numbers in each path
The main things I want to do are:
Find genes that change significantly in both 25C and 37C for each cultivar
Then, we can compare expression patterns to see if they are changing in the same directions
Find genes that are significantly changing only in the 37C condition for each cultivar
Essentially "subtracting" the genes that are also changing at 25C so we are left with uniquely changing 37C genes
Make a heat map of these genes like Figure 13D in Anthony's paper, where each cultivar and temp is on the x-axis (8 sections with 2 time points each = 16 columns to show gene expression levels), and genes are on the y-axis.
This heat map could be made for only the genes significantly changing for all temps/time points/cultivars, and another heat map could be made for the ones only significantly changing in all three thermotolerant lines, but not Heinz
If there are too many genes, we could try to create a "cutoff" with the "Max_PP" section?
Would these things be possible to do? I know Gloria's data did not previously account for temperature, so if there is a better way to visualize the data, I am open to other methods.
If you would like to meet about these things, I am also happy to find a time to talk this over with you. Please let me know what you think.
Thanks so much!
Best,
Kelsey
Attachments
Issue Links
Activity
Tested cedar's 8hr script to ensure all is the same and that we are on the right track.
Performed deseq and got results tables for each variety (3hr 25C vs 3hr 37C)
Tables are stored here: https://drive.google.com/drive/folders/1jeRiJdTZZhF60j7oqgKQRTV3zsgrCkKN
Get gene differentiation for 8hr timepoint using Cedar's script and then my own to ensure consistent results
Ran through Cedar's script, seeing similar results to mine, but seeing more down regulated while I saw more up regulated. Perhaps a flip in one of our designs. Added resulting figures to googles slide presentation.
We should check the top 10 genes according to P adjust and see the pattern cedar way, via your code and at 3 hours.
R
Kelsey wants the various gene lists displayed in my complex upset plot.
The essential ones are:
Unique gene lists (all varieties)
Core genes (shared genes all varieties)
All except Heinz
needs LFC, basemean, basically full deseq output
Write python script to get all these pieces from upset plot and upload to drive
DO FOR 3 HOUR and 8 HOUR
Each list will be it's own file. Tab delimited.
Next steps:
Make the deseq result tables.
repeat for 25C and 37C
Repeat X4 var
Make 1 BIG table saving the log fold change values.
Save tables to the Google drive:
https://drive.google.com/drive/folders/1Vp6US0ItacF0g9m9QR1FqOwXFQB9tf_S?usp=drive_link