User: david.watson

gravatar for david.watson
david.watson10
Reputation:
10
Status:
New User
Location:
Queen Mary University of London
Last seen:
6 months, 4 weeks ago
Joined:
3 years, 7 months ago
Email:
d*******@qmul.ac.uk

Posts by david.watson

<prev • 12 results • page 1 of 2 • next >
0
votes
1
answer
329
views
1
answers
Comment: C: Testing moderated fold changes in DESeq2 version >= 1.16.0
... Hi Michael,   Thanks so much for the prompt and informative reply. I didn't realise that setting betaPrior = TRUE ensures back compatibility when running results(), so that's especially helpful. I recently read the Stephens paper and am very interested to see how local false sign rates may improve a ...
written 20 months ago by david.watson10
1
vote
1
answer
329
views
1
answer
Testing moderated fold changes in DESeq2 version >= 1.16.0
... I'm trying to run a straightforward Wald test in DESeq2 and was a little surprised to see that fold changes and standard errors are no longer shrunken by default. I know I can still compute moderated values using the lfcShrink() function, but I see no straightforward way to: a) calculate the corres ...
deseq2 written 20 months ago by david.watson10 • updated 20 months ago by Michael Love25k
0
votes
1
answer
2.4k
views
1
answers
Comment: C: DESeq2, plotting residuals vs.fitted values
... Follow up question - it appears that sizeFactors for DESeqDataSet objects made with tximport are NULL, even after running the complete DESeq pipeline. This is presumably because the information in assays(dds)[["avgTxLength"]] somehow substitutes for this step. But in that case, how would one go abou ...
written 2.5 years ago by david.watson10
0
votes
1
answer
536
views
1
answers
Comment: C: DESeq2 degrees of freedom
... Got it. Thanks for the prompt reply! ...
written 2.5 years ago by david.watson10
1
vote
1
answer
536
views
1
answer
DESeq2 degrees of freedom
... Is there any simple way to extract the degrees of freedom from a DESeqDataSet object that's been through the complete DESeq pipeline? At first I thought perhaps the empirical Bayes shrinkage augmented the df as it does in limma, but that doesn't appear to be the case judging by the Love et al., 201 ...
deseq2 written 2.5 years ago by david.watson10 • updated 2.5 years ago by Michael Love25k
0
votes
1
answer
837
views
1
answers
Comment: C: Parallelization in DESeq2
... That was another option we've tried. It generates fewer DE genes and the p-value distribution is still a little wonky, suggesting that it maybe fails to capture some of the residual intra-subject correlation, but I understand why you feel it may be more theoretically justified in this case. And ther ...
written 2.8 years ago by david.watson10
0
votes
1
answer
837
views
1
answers
Comment: C: Parallelization in DESeq2
... We have 90 samples, but they are indeed highly correlated since they're from just 10 patients. (Three tissue types observed at three time points each, per subject). We were originally modelling tissue-times independently, but I figured we'd have greater power pooling the data and differentiating tis ...
written 2.8 years ago by david.watson10
0
votes
1
answer
837
views
1
answers
Comment: A: Parallelization in DESeq2
... Hi Michael,   Thanks for the tip about the full argument, I didn't realise DESeq could take a user-supplied design matrix. My model has a large number of coefficients -- 40 if I include all the surrogate variables generated by svaseq without manually fixing n.sv -- and quite a number of genes are ...
written 2.8 years ago by david.watson10
0
votes
1
answer
837
views
1
answer
Parallelization in DESeq2
... Hello,   I know you can set parallel=TRUE within the DESeq and results functions of the DESeq2 package, but I'm wondering if there's any way to parallelize the estimateDispersions and/or nbinomWaldTest functions? Presumably there is since they're doing the heavy lifting within the DESeq function, ...
deseq2 written 2.8 years ago by david.watson10
0
votes
2
answers
665
views
2
answers
Answer: A: Champ.lasso error @ dmr.beta.means[,22:24]
... There's an extra comma in the source code for champ.lasso at line 211 defining dmr.beta.means. You can tell from the error message, where there are commas before and after the column specification. Try deleting the second comma and rerunning.    Cheers, David ...
written 3.5 years ago by david.watson10

Latest awards to david.watson

No awards yet. Soon to come :-)

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 123 users visited in the last hour