From Cuffdiff Output to CummeRbund analysis/visualisation

Question

[Wet&DryImmunology]

Hi,I had problem with using CummeRbund to analyse the output from Cuffdiff.

I was trying to analysis RNA-seq using Cufflinks (I mean the suite).

I got the BAM file after using Tophat and Bowtie for mapping.

Since I am not interested in splicing variants this time, so I skipped using Cufflinks(the program) and Cuffmerge, directly move to Cuffdiff with BAM file, like the workflow highlighted in bright blue here https://wikis.utexas.edu/display/bioiteam/Introduction+to+tophat+cufflinks+workflow

Since I could not successfully install CummeRbund on my server: https://support.bioconductor.org/p/73617/

So I decided to analyse and plot with CummeRbund on my Mac, installed CummeRbund and copy all the output files from my server.

but when I input command:

> cuff<-readCufflinks() # which are supposed to automatically find the default files and begin parsing my data
> cuff # Nothing was done.
CuffSet instance with:
     0 samples
     0 genes
     0 isoforms
     0 TSS
     0 CDS
     0 promoters
     0 splicing
     0 relCDS

When I tried to test the readCufflinks command using the test dataset

> cuff <- readCufflinks(dir=system.file("extdata", package="cummeRbund"))
> cuff    # it seems that it worked.
CuffSet instance with:
     3 samples
     400 genes
     1203 isoforms
     662 TSS
     906 CDS
     1062 promoters
     1986 splicing
     990 relCDS

Then I wonder whether the Cuffdiff was carried out successfully or not, so I checked gene_exp.diff by

cat gene_exp.diff

test_id    gene_id    gene    locus    sample_1    sample_2    status    value_1    value_2    log2(fold_change)    test_stat    p_value    q_value    significant
NM_001001130    NM_001001130    -    chr13:67830198-67857775    q1    q2    OK    5.07206    9.76439    0.944959    1.37033    0.1484    0.869478    no
NM_001001144    NM_001001144    -    chr9:110235791-110287453    q1    q2    OK    7.58925    7.62    0.00583411    0.00620651    0.99445    0.998025    no
NM_001001152    NM_001001152    -    chr13:67355853-67370004    q1    q2    OK    1.04851    1.34624    0.36059    0.373615    0.7031    0.973148    no
NM_001001160    NM_001001160    -    chr6:85419569-85452988    q1    q2    NOTEST    6.48556e-05    8.73562e-05    0

it seems to me that Cuffdiff worked. So I have no clue about what went wrong and how to move on. Can anyone have any suggestions?

If any further information is need to evaluate the problem, please kindly let me know.

Thanks!

Answer 1

[Tky]

Did you copy the CuffDiff results to your current working directory of R?

Or you may set the directory by

cuff <- readCufflinks(dir="/your/cuffdiff/data")

[Wet&DryImmunology]

I copied the all the results to my home directory, and I set the directory by

cuff <- readCufflinks(dir="/Users/shunki")

it did not work.

My colleague noticed that I did not create a specific directory specific for the CuffDiff results (I just copied all the files to the working directory, that might be the problem, so I tried to redo the CuffDiff:

cuffdiff -o diff_out /u32/huang/ucsc_repo_mm9/refGene.gtf pre1.merge.hits_remap.sort.bam \
    tTnaive.merge.hits_remap.sort.bam # specify a "diff_out" directory for all the CuffDiff files exclusively.

and it worked.

so the solution is to set exclusive directory for all the output data as input for readCufflinks, basically as you suggested. Tky.

Thanks!

[Tky]

Cool. Glad you found the solution.