Moderator: Gordon Smyth

gravatar for Gordon Smyth
Gordon Smyth31k
Reputation:
31,010
Status:
Trusted
Location:
Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia
Website:
http://www.statsci.org...
Scholar ID:
Google Scholar Page
Last seen:
7 minutes ago
Joined:
14 years, 7 months ago
Email:
s****@wehi.edu.au

Head of Bioinformatics Division at the Walter and Eliza Hall Institute of Medical Research

Posts by Gordon Smyth

<prev • 2,991 results • page 2 of 300 • next >
0
votes
2
answers
147
views
2
answers
Answer: A: Choosing genes for clustering
... Your question assumes that there is some trick to choosing DE genes, but really there isn't. By definition, the DE analysis is choosing the genes that best separate the two groups. So the best set of DE genes to separate the two sample groups are just the top DE genes. The top gene separates the gro ...
written 6 days ago by Gordon Smyth31k
0
votes
0
answers
108
views
0
answers
Comment: C: Selecting genes that best separate between two groups
... In my opinion there is no good way to conduct a DE analysis of TPM values from Kallisto. TPM values throw away too much information that DE software needs to know. I would be surprised though if that is the entire source of the problem.   ...
written 6 days ago by Gordon Smyth31k
0
votes
0
answers
108
views
0
answers
Comment: C: Selecting genes that best separate between two groups
... If you get lots of DE genes from limma, and you cluster the samples on those genes, then the samples will separate by group. That is always true. There may be other subgroups as well, but certainly the main groups should separate. If this doesn't happen then it follows that you must have done someth ...
written 6 days ago by Gordon Smyth31k
0
votes
2
answers
125
views
2
answers
Answer: A: Treat Method - number of DE genes
... treat() gives more emphasis to genes with larger fold changes, and you can choose whatever lfc seems appropriate depending on how large the changes are in your experiment. In practice, you can choose the lfc cutoff to get the number of DE genes that you want to analyse. Personally though, I would ve ...
written 8 days ago by Gordon Smyth31k
0
votes
2
answers
125
views
2
answers
Comment: C: Treat Method - number of DE genes
... treat() always controls the FDR correctly whatever lfc you choose. Relative to that lfc. The whole idea is that you choose a lfc cutoff that is meaningful to you and appropriate for your dataset. If there was one "right" lfc that was always best then we would set it for you and it wouldn't be an ar ...
written 9 days ago by Gordon Smyth31k
0
votes
2
answers
125
views
2
answers
Comment: C: Treat Method - number of DE genes
... It isn't clear from your post what the problem is. Evidently the software is working correctly. What are you trying to do that you haven't been able to do? You've chosen a very high value for 'lfc', so getting no DE genes from treat() isn't surprising. ...
written 9 days ago by Gordon Smyth31k
0
votes
1
answers
129
views
1
answers
Answer: A: Limma RNAseq comparison with Cuffdiff
... Two things: First you need to define the design matrix as design <- model.matrix(~ classes) The use of "0+" in your formula is out of place and is causing you to test the wrong hypothesis. You are not actually testing for DE between MDK and NEG at all. Second, you need to filter out genes fo ...
written 11 days ago by Gordon Smyth31k
0
votes
2
answers
195
views
2
answers
Comment: C: Limma and batch effect
... Thank you for the extra information. The fact that the unstranded dataset contains only one biological group (not responsive) means that it cannot contribute any information to finding DE genes between responsive and not responsive. So there is no value in combining the two datasets and no need to ...
written 12 days ago by Gordon Smyth31k
0
votes
2
answers
3.0k
views
2
answers
Answer: A: remove microarray batch effects using Limma
... Approach 1 is correct (as I think you already know) and Approach 2 is wrong. Approach 2 is wrong for two reasons. First, the batch effect and condition effects may overlap, and so removeBatchEffect() may have removed some of the condition effect with the batch effect. Second, approach 2 misleads lm ...
written 12 days ago by Gordon Smyth31k
0
votes
2
answers
195
views
2
answers
Answer: A: Limma and batch effect
... To understand removeBatchEffects() you could look at previous answers https://support.bioconductor.org/p/36029 and https://support.bioconductor.org/p/60879 and https://support.bioconductor.org/p/66251 However, removeBatchEffects() is never needed for a DE analysis and I suspect it is just a distrac ...
written 12 days ago by Gordon Smyth31k

Latest awards to Gordon Smyth

Scholar 3 months ago, created an answer that has been accepted. For A: Identification of DEGs through limma analysis
Appreciated 3 months ago, created a post with more than 5 votes. For BH vs BY in p.adjust + FDR interpretation
Popular Question 3 months ago, created a question with more than 1,000 views. For Limma - RNA-Seq DE genes - Quantile normalized log transformed RPKM data
Popular Question 4 months ago, created a question with more than 1,000 views. For limma: paired + multiple comparisons + technical replication?
Scholar 4 months ago, created an answer that has been accepted. For A: limma; Only fit list of certain genes/ filter eSet
Popular Question 4 months ago, created a question with more than 1,000 views. For total count filter cutoff (edgeR)
Teacher 4 months ago, created an answer with at least 3 up-votes. For A: Subsetting ELists and Applying sva to Iterative Differential Gene Expression Tes
Teacher 5 months ago, created an answer with at least 3 up-votes. For A: Subsetting ELists and Applying sva to Iterative Differential Gene Expression Tes
Popular Question 5 months ago, created a question with more than 1,000 views. For Filtering is not recommended with LIMMA?
Good Answer 5 months ago, created an answer that was upvoted at least 5 times. For A: Confidence intervals on edgeR logFC
Scholar 5 months ago, created an answer that has been accepted. For A: limma; Only fit list of certain genes/ filter eSet
Teacher 5 months ago, created an answer with at least 3 up-votes. For A: limma; Only fit list of certain genes/ filter eSet
Appreciated 5 months ago, created a post with more than 5 votes. For BH vs BY in p.adjust + FDR interpretation
Scholar 5 months ago, created an answer that has been accepted. For A: limma; Only fit list of certain genes/ filter eSet
Popular Question 5 months ago, created a question with more than 1,000 views. For total count filter cutoff (edgeR)
Commentator 5 months ago, created a comment with at least 3 up-votes. For C: How did the edgeR authors compute Figure 2 (genewise deviance statistics?)
Commentator 5 months ago, created a comment with at least 3 up-votes. For C: How did the edgeR authors compute Figure 2 (genewise deviance statistics?)
Teacher 6 months ago, created an answer with at least 3 up-votes. For A: limma; Only fit list of certain genes/ filter eSet
Scholar 6 months ago, created an answer that has been accepted. For A: limma; Only fit list of certain genes/ filter eSet
Teacher 6 months ago, created an answer with at least 3 up-votes. For A: seeking a paper or journal review about repeated multiple testing for Biological
Teacher 6 months ago, created an answer with at least 3 up-votes. For A: limma; Only fit list of certain genes/ filter eSet
Teacher 6 months ago, created an answer with at least 3 up-votes. For A: limma; Only fit list of certain genes/ filter eSet
Scholar 6 months ago, created an answer that has been accepted. For A: limma; Only fit list of certain genes/ filter eSet
Scholar 6 months ago, created an answer that has been accepted. For A: limma; Only fit list of certain genes/ filter eSet

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.2.0
Traffic: 102 users visited in the last hour