tophat:all p-value are greater than 5e-05 ??

[SpreeFu]

Hello all,

when I plot my data by cummeRbund with the function
> v<-csVolcanoMatrix(genes(cuff))
> v

the plot showed all p-value are greater than 5e-05.
But I could not find how I get this. Please help me, Thanks a lot!

[dpryan]

From the CummeRbund help page:

> gene.diff<-diffData(genes(cuff))
> head(gene.diff)
 

      gene_id sample_1 sample_2 status   value_1     value_2
1 XLOC_000001      iPS     hESC NOTEST  20.21750 3.47386e-01
2 XLOC_000002      iPS     hESC NOTEST   0.00000 0.00000e+00
3 XLOC_000003      iPS     hESC NOTEST   0.00000 0.00000e+00
4 XLOC_000004      iPS     hESC     OK   0.00000 6.97259e+05
5 XLOC_000005      iPS     hESC     OK 355.82300 6.96704e+02
6 XLOC_000006      iPS     hESC NOTEST   1.51396 0.00000e+00
  log2_fold_change     test_stat    p_value    q_value
1     -5.86292e+00   7.13525e-01 0.47552100 1.00000000
2      0.00000e+00   0.00000e+00 1.00000000 1.00000000
3      0.00000e+00   0.00000e+00 1.00000000 1.00000000
4     1.79769e+308  1.79769e+308 0.00857693 0.02109120
5      9.69385e-01  -2.98373e+00 0.00284757 0.00840284
6    -1.79769e+308 -1.79769e+308 0.24382400 1.00000000
  significant
1          no
2          no
3          no
4         yes
5         yes
6          no

So gene.diff$q_value should hold the values to use.

[SpreeFu]

From the CummeRbund help page:

> gene.diff<-diffData(genes(cuff))
> head(gene.diff)
 

      gene_id sample_1 sample_2 status   value_1     value_2
1 XLOC_000001      iPS     hESC NOTEST  20.21750 3.47386e-01
2 XLOC_000002      iPS     hESC NOTEST   0.00000 0.00000e+00
3 XLOC_000003      iPS     hESC NOTEST   0.00000 0.00000e+00
4 XLOC_000004      iPS     hESC     OK   0.00000 6.97259e+05
5 XLOC_000005      iPS     hESC     OK 355.82300 6.96704e+02
6 XLOC_000006      iPS     hESC NOTEST   1.51396 0.00000e+00
  log2_fold_change     test_stat    p_value    q_value
1     -5.86292e+00   7.13525e-01 0.47552100 1.00000000
2      0.00000e+00   0.00000e+00 1.00000000 1.00000000
3      0.00000e+00   0.00000e+00 1.00000000 1.00000000
4     1.79769e+308  1.79769e+308 0.00857693 0.02109120
5      9.69385e-01  -2.98373e+00 0.00284757 0.00840284
6    -1.79769e+308 -1.79769e+308 0.24382400 1.00000000
  significant
1          no
2          no
3          no
4         yes
5         yes
6          no

So gene.diff$q_value should hold the values to use.

Thanks dpryan! But I'm sorry, I did not get the meaning 'gene.diff$q_value should hold the values to use', could you explain more details please?

[dpryan]

Hi spreefu,

My understanding is that you're trying to extract the p-values (or probably more useful, the q-value) from your dataset. These are available using the "diffData" function which returns a data.frame (the top portion of example output was provided in my earlier reply). This data.frame contains the p and q-values that you seem to be looking for. If you just want to double check that the lowest p-value is > 5e-5, then just:

gene.diff<-diffData(genes(cuff))
min(gene.diff$p_value)

The "$p_value" just says, "just look at the column labeled p_value". You'll find a basic fluency in R to be very helpful in bioinformatics, so I would recommend going through a few tutorials (or taking a class if there's one offered locally).

[SpreeFu]

Hi spreefu,

My understanding is that you're trying to extract the p-values (or probably more useful, the q-value) from your dataset. These are available using the "diffData" function which returns a data.frame (the top portion of example output was provided in my earlier reply). This data.frame contains the p and q-values that you seem to be looking for. If you just want to double check that the lowest p-value is > 5e-5, then just:

gene.diff<-diffData(genes(cuff))
min(gene.diff$p_value)

The "$p_value" just says, "just look at the column labeled p_value". You'll find a basic fluency in R to be very helpful in bioinformatics, so I would recommend going through a few tutorials (or taking a class if there's one offered locally).

Hi dpryan,
Thanks for your suggestion! I know if I want to find the significant genes, I should take q-value.
But I still have the question about those p-values in my data. Because p-values of many genes are 5e-5, it seems 5e-5 is a limit here. Then when I generate the volcano-plot, you can see my picture attach in the first post, it looks the volcano was cut by a line.
If you have look in cummeRbund manual here:

http://compbio.mit.edu/cummeRbund/manual_2_0.html#tth_sEc6

the p-value in their volcano-plot do not limit by 5e-5. This is what I wondered.
Thanks!

[SpreeFu]

Hi dpryan,

Thanks for your help! Sorry for confusing you because of I did not describe my question clearly at first.
For the question here 'But I still have the question about those p-values in my data. Because p-values of many genes are 5e-5, it seems 5e-5 is a limit here. Then when I generate the volcano-plot, you can see my picture attached in the first post, it looks the volcano was cut by a line.
If you have look in cummeRbund manual here:

http://compbio.mit.edu/cummeRbund/ma....html#tth_sEc6

the p-value in their volcano-plot do not limit by 5e-5. This is what I wondered.' Do you have any idea?

[dpryan]

The p-values come from cuffdiff, but unfortunately I don't know enough about its inner-workings to give you a good answer to that.

[SpreeFu]

The p-values come from cuffdiff, but unfortunately I don't know enough about its inner-workings to give you a good answer to that.

Anyway, still thanks a lot! Schönes Wochenende!

[yzzhang]

In the thread http://http://seqanswers.com/forums/...ad.php?t=30705
just mentioned that, hope it helps

[SpreeFu]

hi yzzhang, thanks for your help!