0

I have two datasets one is log10 transformed and the other is RMA normalized.

DATAFRAME 1(protdata)

 dim(protdata) [1] 437   7
      Locus      COBL9   WER    CORTEX   SCR     WOL    PET111
  1  AT1G01090  4.3035  4.3975  4.4620  4.3879  4.2620  4.2045
  2  AT1G02780  4.7852  4.6865  4.7801  5.0038  4.9457  4.9515
  3  AT1G04040  4.5854  4.1787  3.4836  3.4918  4.2366  3.5197
  4  AT1G04270  4.3578  4.4281  4.3348  4.5680  4.4208  4.4056
  5  AT1G04410  4.9808  4.9913  5.2186  5.3315  5.2178  5.4524
  6  AT1G04430  4.2382  4.3564  4.3535  4.3056  4.0263  3.9485
  7  AT1G04480  4.5462  4.4302  4.4987  4.8039  4.5807  4.4876

DATAFRAME 2 (transdata)

 dim(transdata) [1] 22810    77
       Locus      probes     COBL9    WER   CORTEX  SCR     
 1   AT1G01090   244901_at   4.7852 4.6865  4.7801  5.0038  
 2   AT1G02780   244902_at   4.5854 4.1787  3.4836 3.4918                          
 3   AT1G04040   244903_at   4.3578 4.4281  4.3348  4.5680              
 4   AT1G04570   244904_at   4.9808 4.9913  5.2186  5.3315                  
 5   AT1G04610   244905_at   4.2382 4.3564  4.3535  4.3056                  
 6   AT1G02430   244906_at   4.5462 4.4302  4.4987  4.8039

I want to merge the two dataframes based on the Locus ids and used

 matchin<-merge(transdata,protdata,by.x="Locus",by.y="Locus")

But I get a different dataframe which is as follows: In the above example I have two matching ids in both frames but result is :

      Locus      probes     COBL9    WER    CORTEX  SCR     
 1   AT1G01090   244901_at  14.7852 12.6865 13.7801 12.0038     
 2   AT1G02780   244902_at  14.5854 13.1787 12.4836 13.4918 
             COBL9   WER    CORTEX   SCR     WOL    PET111
            4.3035  4.3975  4.4620  4.3879  4.2620  4.2045
            4.7852  4.6865  4.7801  5.0038  4.9457  4.9515

Though it is a dataframe containing unique locus ids the intensity values from the transdata alone has changed.

Community
  • 1
  • 1
user1805343
  • 127
  • 1
  • 2
  • 11

1 Answers1

1

Works for me. Clear your workspace and start a new. Make sure you're using the right objects.

> merge(transdata,protdata, by = "Locus")
      Locus    probes COBL9.x  WER.x CORTEX.x  SCR.x COBL9.y  WER.y CORTEX.y  SCR.y    WOL PET111
1 AT1G01090 244901_at  4.7852 4.6865   4.7801 5.0038  4.3035 4.3975   4.4620 4.3879 4.2620 4.2045
2 AT1G02780 244902_at  4.5854 4.1787   3.4836 3.4918  4.7852 4.6865   4.7801 5.0038 4.9457 4.9515
3 AT1G04040 244903_at  4.3578 4.4281   4.3348 4.5680  4.5854 4.1787   3.4836 3.4918 4.2366 3.5197
Roman Luštrik
  • 69,533
  • 24
  • 154
  • 197
  • I tried again and when used on the whole datset the intesityvalues from tansdata alone were like tripled to the original intesity while merging – user1805343 Feb 26 '13 at 09:14
  • 1
    @user1805343, a fully reproducible example would help, I bet. – BenBarnes Feb 26 '13 at 09:42
  • https://gist.github.com/anonymous/5037623 https://gist.github.com/anonymous/5037633 are the transdata and protdata respectively – user1805343 Feb 26 '13 at 10:52
  • 1
    @user1805343 Would it be possible to make a mock example that demonstrates the behavior you're experiencing? Here are a few tips on how to do that: http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example There's little guarantee that github data will be available in the future, rendering this question and potential answer less useful. – Roman Luštrik Feb 26 '13 at 11:15
  • Its not possible with the full data as it is too huge and a subset of it not having common things – user1805343 Feb 26 '13 at 12:06