Cummerbund Isn'T Managing Cuffdiff Database

Question

[caworms]

Heya,

I'm trying to output cuffdiff data into cummeRbund, but it appears to be having issues with connection to a database (below). Does anyone have any suggestions as to how I can fix this? Thanks!

cuffdata <- readCufflinks('/home/sbica1/tophat/finaltrans/tophat/0125diffout')

Creating database /home/sbica1/tophat/finaltrans/tophat/0125diffout/cuffData.db Reading /home/sbica1/tophat/finaltrans/tophat/0125diffout/genes.fpkmtracking Checking samples table... Populating samples table... Writing genes table Reshaping geneData table Recasting Writing geneData table Reading /home/sbica1/tophat/finaltrans/tophat/0125diffout/gene_exp.diff Error in function (classes, fdef, mtable) : unable to find an inherited method for function "make.db.names", for signature "SQLiteConnection", "integer"

csVolcano(genes(cuff_data),'0',36)

Error in .local(object, x, y, features, ...) : One or more values of 'x' or 'y' are not valid sample names!

[Michael]

did you install RSQLite?

[caworms]

Yup. (taking up 15 characters)

[John St. John]

Did you try running it twice in a row? Kinda stupid thing to try, but it actually worked for a problem I was having with cummeRbund.

[caworms]

Hi John St. John- I did, but to no avail.

Also- thanks for your suggestion to altering hits.cpp prior to compilation of cufflinks, it sorted out another issue I had.

Answer 1

[Fabio Marroni]

Just to summarize things. Please, correct me if I miss something. In the gene_exp.diff file (and in all the .diff files) there are two columns called sample_1 and sample_2, respectively. These are the sample names (for @mad.cichlids). These columns SHOULD NOT

1) assume the "T" value (@Conana experience, plus confirmation by authors),

2) assume numeric values (according to @caworms and my experience, plus confirmation by authors).

If any of this happens, the corresponding errors mentioned error will be issued.

1) unable to find an inherited method for function "make.db.names", for signature "SQLiteConnection", "logical"

2) unable to find an inherited method for function "make.db.names", for signature "SQLiteConnection", "integer"

Finally, I found a suggestion by Loyal, to avoid names that start with a number and (just to be safe) I suggest to stick to that.

Answer 2

[Conan]

Hey caworms, I have an similar issue of cummeRbund and hope anyone can help.

> library(cummeRbund)     
> cuff<-readCufflinks("cuffdata_120417")

Creating database cuffdata120417/cuffData.db
Reading cuffdata120417/genes.fpkmtracking
Checking samples table...
Populating samples table...
Writing genes table
Reshaping geneData table
Recasting
Writing geneData table
Reading cuffdata120417/gene_exp.diff
Error in function (classes, fdef, mtable) :
unable to find an inherited method for function "make.db.names", for signature "SQLiteConnection", "logical"

That's interesting because it has already a cuff.db there but I'm curious about whether this database is fully built. P.S. It seems like there's something wrong wiz RSQLite package but I'm not that familiar in using R Packages. :(

[caworms]

Heya, R has some pretty decent support forums, so I'll try there and let you know what I find...

[Conan]

Hi there, I've sent email to the author of cummeRbund Loyal and my question solved. The reason why cummeRbund stuck was "I named the sample column 'T' and 'N'" and R deem the "T" as "True" so it reports the logical error. Hope my experience helps and I recommend you report questions to [e-mail] directly.

[caworms]

Heya Conan, I too, contacted Loyal about it in the end. Pretty much the same thing- I named mine after the doses and R doesn't like numbers as names. Bloody R...

Cheers,

Craig

[mad.cichlids]

Hi, Craig I ran into the same problem. What do you mean sample name? Is it the original .fasta file name or something else? Thank you a lot!