## User: jshouse

jshouse10
Reputation:
10
Status:
New User
Location:
Last seen:
2 months, 3 weeks ago
Joined:
3 years, 4 months ago
Email:
j******@yahoo.com

#### Posts by jshouse

<prev • 25 results • page 1 of 3 • next >
2
238
views
2
... I went to estimate cell counts from a 450k RGSet with  estimateCellCounts(RGSet, referencePlatform = c("IlluminaHumanMethylation450k"))  I got the following:  > cellCounts <- estimateCellCounts(RGSet, referencePlatform = c("IlluminaHumanMethylation450k")) [estimateCellCounts] Co ...
written 3 months ago by jshouse10 • updated 4 weeks ago by lucy0
1
352
views
1
... First, this is a rather large experiment. Roughly 3000 features by 1300 samples. Creating the dds` object takes 12.5 hours with 20 workers at 3.1 ghz, and resultsNames(dds)  consists of 487 items.  > resultsNames(dds)[1:8] [1] "Intercept" "group_00 ...
written 20 months ago by jshouse10 • updated 20 months ago by ellascottgm0
1
644
views
1
Comment: C: DESeq2 MA plot look different
... Thanks Michael, I didn't catch that. Sorry for wasting your time. I will look there first next time.   ...
written 2.0 years ago by jshouse10
1
644
views
1
... In the past, (DESeq2 versions < 1.1.4), MA plots from DESeq2 looked like the following (formatting modified by me, but you get the idea). MA plots since 1.14 look different. Is there any reason to be alarmed at how the distribution of l2fc's is dramatically different in its randomness at low c ...
written 2.0 years ago by jshouse10 • updated 2.0 years ago by Michael Love25k
1
1.4k
views
1
Comment: C: DESEQ2 linear model of dose
... My concern is the coefficient for TreatmentB:dose will use count data from treatmentA: doses 1,2,3. Or to put another way, the model as written will model 6 doses of (0) with 1 dose each of 1,2,3 for both TreatmentA:dose and TreatmentB:dose. This may be ignorance on my part regarding the model matri ...
written 3.0 years ago by jshouse10
1
1.4k
views
1
Comment: C: DESEQ2 linear model of dose
... For treatment A, wont that design assign B treatment values to a zero dose? Likewise for B with regard to A. ...
written 3.0 years ago by jshouse10
1
1.4k
views
1
Comment: C: DESEQ2 linear model of dose
... Thanks Michael and Simon.  My data are structured as follows, where A and B share the zero dose from V (vehicle).  > coldata <- data.frame(treatment=factor(rep(c("A","B","V"),each=3)), + dose=c(1:3,1:3,0,0,0)) > coldata treatment dose 1 A 1 2 ...
written 3.0 years ago by jshouse10
1
1.4k
views
1
Comment: C: DESEQ2 linear model of dose
... Is the full rank error because there exists only 1 dose per non-vehicle treatment?  ...
written 3.0 years ago by jshouse10
1
1.4k
views
1
Comment: C: DESEQ2 linear model of dose
... Simon, I am sorry that this is taking so much of your time, and I appreciate you sharing it. Using ~ Treatment ​* logdose returns a full rank error.  I'll try to explain more clearly the experimental design by pasting example below.  These two treatments share the same 12 vehicle controls. On purpos ...
written 3.0 years ago by jshouse10 • updated 3.0 years ago by Michael Love25k
1
1.4k
views
1
Comment: C: DESEQ2 linear model of dose
... Simon, I was rearranging the code when I got stuck.  The data are arranged as posted in the first box of the original question. When design ~ Group, this worked fine as I was using contrasts to pull out Treatment from Vehicle comparisons.  The dataset actually has 73 different treatments structured ...
written 3.0 years ago by jshouse10

#### Latest awards to jshouse

Popular Question 15 months ago, created a question with more than 1,000 views. For DESEQ2 linear model of dose

Content
Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.