Details
-
Type:
Task
-
Status: Closed (View Workflow)
-
Priority:
Major
-
Resolution: Done
-
Affects Version/s: None
-
Fix Version/s: None
-
Labels:None
-
Story Points:3
-
Sprint:Fall 2 2023 Sep 17, Fall 3 2023 Oct 2, Fall 4 2023 Oct 16
Description
Purpose: Write a first draft of an EB-Seq script to analyze the time course data.
Why? Because comparing time points violates an assumption of independence, a few tools have been developed to address this and improve the RNA-seq comparisons on time dependent studies.
A few examples:
- mfuzz
- PAL-D
- EB-Seq
I tested out PAL-D last January and it was a horrible experience.
Rasha ran mfuzz with success.
EB-Seq has been run by me and Liz in the past.
Liz's details on Eb0seq can be found here:
_Original details from Liz Cooper
“EBSeq script on my GitHub page associated with the sweet sorghum paper (in case you decide to use that package): “
Rio/EBSeqHMM_Cluster.R at master · eacooper400/Rio · GitHub
Liz_
In R:
The R package should be acquired like so:
-
-
- Installing Required Packages (only done once)
#source("https://bioconductor.org/biocLite.R")
BiocManager::install("EBSeq")
BiocManager::install("EBSeqHMM")
install.packages("blockmodeling")
- Installing Required Packages (only done once)
-
-
-
- Load Required Pacakges
library(EBSeq)
library(EBSeqHMM)
- Load Required Pacakges
-
Attachments
Issue Links
- is blocked by
-
-
- Closed
-
Activity
For R script, see comments below....... muday-144-SL4_counts-salmon.txt
Instructions for review:
- Download the attached R script. And run it on the attached Salmon counts file.
- Open the script in Rstudio and try to run it. A few libraries will need to be installed possibly depending on your environment.
- You will need to change the locations on lines 20 and 26 to your folder location.
- Takes many minutes to run.
- The end result should be a collection of result files that look similar in size to these:
rw-rr-@ 1 robreid staff 63100 Oct 19 10:24 a28gc_ebseqGeneCalls-SL4.txt
rw-rr-@ 1 robreid staff 63544 Oct 19 10:24 a34gc_ebseqGeneCalls-SL4.txt
rw-rr-@ 1 robreid staff 49124 Oct 19 10:24 v28gc_ebseqGeneCalls-SL4.txt
rw-rr-@ 1 robreid staff 49817 Oct 19 10:24 v34gc_ebseqGeneCalls-SL4.txt
rw-rr-@ 1 robreid staff 40735 Oct 19 10:24 f28gc_ebseqGeneCalls-SL4.txt
rw-rr-@ 1 robreid staff 42402 Oct 19 10:24 f34gc_ebseqGeneCalls-SL4.txt
rw-rr-@ 1 robreid staff 483 Oct 19 10:24 a28gc_numIneachPath-SL4.txt
rw-rr-@ 1 robreid staff 484 Oct 19 10:24 a34gc_numIneachPath-SL4.txt
rw-rr-@ 1 robreid staff 482 Oct 19 10:24 v28gc_numIneachPath-SL4.txt
rw-rr-@ 1 robreid staff 483 Oct 19 10:24 v34gc_numIneachPath-SL4.txt
rw-rr-@ 1 robreid staff 481 Oct 19 10:24 f28gc_numIneachPath-SL4.txt
rw-rr-@ 1 robreid staff 482 Oct 19 10:24 f34gc_numIneachPath-SL4.txt
Take a peak at a result file or 2 and see if it is intuitive!
Review:
- The script runs good in RStudio and ran into no issues.
- I did have issues with the output files. There is a third column after "Max_PP" that is not named and is full of numbers. I am unsure what those numbers represent.
- I also try to open the txt files in excel and they do not visually transfer correctly. The strings are mismatched and do not separate correctly.
- Might be beneficial to create a data frame object for each result and make sure the columns save correctly with column names.
Thanks!
Ebseq-hmm_mudayTimeCourse-Version3.R
Updated version that handles the headers and keeps things in line for better exporting to Excel.
Molly can you run this version?
Review:
- The EBSeq output txt files are fixed now and column names are fixed.
- The numIneachPath txt files still have some formatting issues but that's only if you try to open in excel. These files are readable as just txt files so I see no reason for them needing to be readable in excel also.
Moving ticket to done!
Here is journal article discussing rna-seq statistical methods. They discuss EB-SEQ.
And they mention that is performs well when n=3, as in this project!
journal.pone.0264246.pdf