Entering edit mode
4.1 years ago
eridanus
▴
40
Hello. I am trying to use tximeta package. I have quant.sf files from salmon. my code is the following one:
dir <- "/home/quant_sf"
names<-c("28","29","30","31","32","33","34","35","36","37","38","39")
files <-file.path(dir, names,"quant.sf")
files
file.exists(files)
coldata <- data.frame(files,names=names, stringsAsFactors=FALSE)
coldata
se <- tximeta(coldata)
I am getting the following error:
Error: lexical error: invalid char in json text.
/home/quant_sf
(right here) ------^
Could someone help me? Thank you so much in advance!!
I think I see the problem with your code BUT even when corrected I didn't get the desired output from tximeta
I ended up just using
quantmergeto combine my salmon outputs and imported that as a count matrix because it was way easier.Then use tximport. It does the exact same except it does not add the metadata which are (imho) mostly not necessary for a standard analysis. tximeta seems a bit overloaded to me right now, and this annoying json dependency is one of the factors it is currently unattractive compared to tximport which is super easy to use. Also, quantmerge does not aggregate to the gene level so this is not really a workaround afai can see.
Now I am trying to use tximport. However I am having a hard time creating tx2gene. I have used gencode.v33.transcripts.fa.gz and Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz for indexing. Any thoughts? Thank you!
could you tell me the error in my code? thanks!!
quant_sf is a folder containing 12 more folders -named as names vector- (each folder contains 1 quant.sf file). So for example: /home/quant_sf/28/quant.sf is the full path.
I remember that I also had that error and was annoyed by it, and simply used tximport. Suggest to ask this over at
support.bioconductor.orgso the developer can have a look. Please include output ofsessionInfo()if you do.thank you so much for your reply!i will try tximport..