## User: Belinda Phipson

Reputation:
130
Status:
Trusted
Location:
Last seen:
5 years, 11 months ago
Joined:
7 years, 3 months ago
Email:
p******@wehi.edu.au

#### Posts by Belinda Phipson

<prev • 12 results • page 1 of 2 • next >
1
576
views
1
... Hi Christian You can use points() to add coloured dots on the figure after you have run the genas command. The x and y axes are from fit$coeff. For example, you can identify your significant genes using the match command (or from decideTests() etc.): m<-match(sig_genes,fit$genes$Symbol) genas(f ... written 5.9 years ago by Belinda Phipson130 1 answer 1.2k views 1 answers ... Hi Michelle The function propexpr() in the limma package will estimate the proportion of expressed probes in each array, as long as there are some negative controls present in your data. You can also estimate the number of probes that are not differentially expressed between WT and knock-out using ... written 7.2 years ago by Belinda Phipson130 1 answer 4.8k views 1 answers ... Hi Jack I would fit my design matrix like this: > design<-model.matrix(~A+B) and because there are only two levels for A it is treated as a factor, but B is treated as a continuous variable. > design (Intercept) A B 1 1 1 25 2 1 0 35 3 1 0 28 4 ... written 7.2 years ago by Belinda Phipson130 1 answer 1.1k views 1 answers ... Hi Steven A common problem with small sample sizes! There are some things you can try: 1) You can try using a function called combat() in the sva package to remove the cell line effect. 2) You can have a look at what fit$df.prior is giving you. A larger value will result in more significant differe ...
written 7.2 years ago by Belinda Phipson130
1
527
views
1
... Hi Jack I think you have misunderstood what lmFit does. lmFit takes the data object/matrix, call it x, and fits a user-specified design matrix to it, call it design. i.e. > fit <- lmFit(x,design) >From your message I don't understand what format your data is in. However, if you have two ...
written 7.2 years ago by Belinda Phipson130
2
749
views
2
... Hi Gowthaman Your output looks fine. What is more important is that library size is taken into account as an offset later on when you fit the glm. See help(glmFit). Cheers, Belinda -----Original Message----- From: bioconductor-bounces@r-project.org [mailto:bioconductor- bounces@r-project.org] On ...
written 7.3 years ago by Belinda Phipson130
1
1.1k
views
1
... Hi Steven You could just include cell line in your linear model rather than using duplicateCorrelation(). > design <- model.matrix(~factor(targets$cellline)+factor(targets$fenotype)) > fit <- lmFit(eset,design) > fit <- eBayes(fit) This will test R vs S taking into account cell ...
written 7.3 years ago by Belinda Phipson130
1
431
views
1
... Hi Mei Check the names of your data object: > names(data) to figure out where the normalized data is and then use the > write.csv(data\$...,file="norm.csv") which can write matrices or data frames to a file which can be opened in excel. Cheers, Belinda -----Original Message----- From: biocon ...
written 7.3 years ago by Belinda Phipson130
1
406
views
1
... Hi Murali There is no automatic filtering in the limma package, if that is what you are asking. There are different approaches one could take to filter out genes - often lowly expressed genes are filtered out, or in the case of Illumina microarrays, you can filter out probes with low detection valu ...
written 7.3 years ago by Belinda Phipson130
1
777
views
1
... Hi Neta On page 16 and 17 of the limma vignette (25 March 2012) there is an explanation about column names in the data files. You should check what the header names of your Agilent files are and then change the read.maimages() command accordingly. For example: > RG <- read.maimages(files, + ...
written 7.3 years ago by Belinda Phipson130

#### Latest awards to Belinda Phipson

No awards yet. Soon to come :-)

Content
Help
Access

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