User: Gavin Kelly

gravatar for Gavin Kelly
Gavin Kelly230
Reputation:
230
Status:
Trusted
Location:
United Kingdom / London / Francis Crick Institute
Twitter:
gavinpaulkelly
Scholar ID:
Google Scholar Page
Last seen:
18 minutes ago
Joined:
2 years, 9 months ago
Email:
g*************@gmail.com

Statistician in the Bioinformatics & Biostatistics group at the Francis Crick institute

Posts by Gavin Kelly

<prev • 61 results • page 1 of 7 • next >
0
votes
2
answers
136
views
2
answers
Answer: A: deseq2 complex time course data with out replicates
... While completely agreeing with Michael's answer on the impossibility of testing pairs of timepoints, I'm wondering if you could take some prior expectations into account.  If you're expecting WT and KO to behave similarly at certain timepoints (may be the 10min placebo), then you could label those s ...
written 2 days ago by Gavin Kelly230
0
votes
2
answers
64
views
2
answers
Answer: A: [DESeq2] time-course Experimental design
... Still a bit confused about your use of LRT in the 'cheaper' design - it does not look at 'all... timepoints versus the 0h' - it's looking to see if the null of all timepoints being the same holds.  So an LRT could feasibly come out as significant if the end timepoint was different from all previous ...
written 2 days ago by Gavin Kelly230
0
votes
2
answers
64
views
2
answers
Answer: A: [DESeq2] time-course Experimental design
... Sounds like you'd just need a single factor to model the experiment, where if U corresponds to 0hours, and e.g. T has 2hours, 4hours 16hours, then you'd simply label the replicates with a single factor, with levels  0h, 2h, 4h, 16h etc.   You obviously won't be able to infer anything about how the u ...
written 3 days ago by Gavin Kelly230
0
votes
1
answers
83
views
1
answers
Comment: C: DESeq2 resultsNames(dds) not giving the right comparisons?.
... Yes, your interpretation of out.21w is correct.  The DESeq() doesn't actually do any testing, it's just fitting the model, so no comparisons are done at that stage.  It's only when 'results' is called that the tests are run. I guess the reason for resultsNames is to give beginners, probably most of ...
written 5 days ago by Gavin Kelly230
0
votes
1
answers
83
views
1
answers
Answer: A: DESeq2 resultsNames(dds) not giving the right comparisons...but results(dds, con
... It looks like everything is working as intended.  The resultsNames function gives you a list of comparisons you can do through the shortcut of specifying a name rather than a contrast to the results function, and compares everything to a baseline condition (in your case 16wC, as that is the first al ...
written 6 days ago by Gavin Kelly230
0
votes
2
answers
64
views
2
answers
Answer: A: DESeq2 using interaction terms or grouping variable
... Depending on which version of DESeq2 you're using, you can get different methods of moderating the fold-change estimates.  It defaulted to shrinking the estimates towards a global (across genes) fold-change if there was no interaction term present.  But if you used an interaction term, then this beh ...
written 9 days ago by Gavin Kelly230
0
votes
1
answers
88
views
1
answers
Comment: C: DESeq2 time series tissue specific differences over time
... "summands should correspond to things in the resultsNames"  just means I was too lazy to write TissueB.Time5, and just wrote B.5 in the table, that you should be able to work out which I really meant, and that all of the individual subcomponents should correspond to something you see in resultsNames ...
written 10 days ago by Gavin Kelly230
0
votes
1
answers
88
views
1
answers
Answer: A: DESeq2 time series tissue specific differences over time
... I always like to tabulate group-averages in terms of their component estimates when I'm talking through possible comparisons   1 5 10 A I I+'5' I + 10 B I + B I+B + 5 + B.5 I+ B + 10 +B.10 (the summands all correspond to things in results ...
written 22 days ago by Gavin Kelly230
0
votes
1
answers
98
views
1
answers
Comment: C: Inverse logFC on DESeq2 Rstudio
... In your contrast=c("condition", "CTL", "HT") you're saying that you're looking at fold changes relative to "HT" - the second element in the character vector is the numerator, and the third is the level that will be denominator (or 'be subtracted' on the log scale).  Guessing that "CTL" means control ...
written 7 weeks ago by Gavin Kelly230
0
votes
1
answers
94
views
1
answers
Comment: C: DESeq2 on 2 genotypes versus 1 control
... If, by common, you mean having an identical difference (upto noise) in both genotypes when compared to control, then you could create another variable along the lines of condition2 <- ifelse(condition=="control", "control", "common_genotype") which would 'pool' the two genotypes, but this is prob ...
written 7 weeks ago by Gavin Kelly230

Latest awards to Gavin Kelly

Scholar 4 months ago, created an answer that has been accepted. For A: DESeq2 3-fatcor design and different interaction terms
Teacher 4 months ago, created an answer with at least 3 up-votes. For A: extracting the genes associated with the clusters
Autobiographer 4 months ago, has more than 80 characters in the information field of the user's profile.
Scholar 20 months ago, created an answer that has been accepted. For A: DESeq2 3-fatcor design and different interaction terms

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.2.0
Traffic: 242 users visited in the last hour