How to extract Common Genes from 2 spreadsheets

byou678 · 10-03-2011, 08:53 AM

Now I have two Excel spreadsheets, parts of them are as below:

Spreadsheet1

Gene Change Padj
1 212 0.0006
2 99 0.09
3 15 0.7
4 59 0.456
5 418 0.00001
6 566 0

Spreadsheet2

Gene Statistic FDR
1 28 0.0005
2 17 0.007
3 3 0.81
4 8 0.28
5 48 0.002
6 111 0

Using Excel, how can I create one new spreadsheet that only includes the
Genes with both Padj and FDR are less than 0.05?

And how can I do this in any other statsitical software, Like R, SPSS?
Thanks very much for any idea for any question!

ECO · 10-03-2011, 09:51 AM

This is cake in R...

Depending on how large the datasets are, read them into R. The example below (using read.table) is just for using the clipboard in MacOSX (highlight in excel, copy, then run that line). The equiv in windows is read.table("clipboard",sep='\t',header=TRUE).

Code:

> df1 <- read.table(pipe("pbpaste"),header=TRUE)
> df1
  Gene Change    Padj
1    1    212 0.00060
2    2     99 0.09000
3    3     15 0.70000
4    4     59 0.45600
5    5    418 0.00001
6    6    566 0.00000
> df2 <- read.table(pipe("pbpaste"),header=TRUE)
> df2
  Gene Statistic    FDR
1    1        28 0.0005
2    2        17 0.0070
3    3         3 0.8100
4    4         8 0.2800
5    5        48 0.0020
6    6       111 0.0000
> df3 <- merge(df1,df2)
> df3
  Gene Change    Padj Statistic    FDR
1    1    212 0.00060        28 0.0005
2    2     99 0.09000        17 0.0070
3    3     15 0.70000         3 0.8100
4    4     59 0.45600         8 0.2800
5    5    418 0.00001        48 0.0020
6    6    566 0.00000       111 0.0000
> df3[df3$Padj< 0.05 & df3$FDR < 0.05,]
  Gene Change  Padj Statistic   FDR
1    1    212 6e-04        28 5e-04
5    5    418 1e-05        48 2e-03
6    6    566 0e+00       111 0e+00
> write.csv(df3[df3$Padj< 0.05 & df3$FDR < 0.05,])
"","Gene","Change","Padj","Statistic","FDR"
"1",1,212,6e-04,28,5e-04
"5",5,418,1e-05,48,0.002
"6",6,566,0,111,0

byou678 · 10-03-2011, 12:57 PM

Thanks ECO, I run the similar codes in R, too. And I can get the result like yours Just based on the Spreadsheets showed in this thread. Sorry for not telling more info about my data:But I have more Genes(37970) in Spreadsheet2 than Spreadsheet1 (17980). It supposed to have 17980 Genes left after merging [the 19990(37970-17980) Genes which only in Spreadsheet2 will not appear in the merged sheets]. Unfortunately, it messed up after merging. Any idea will be greatly appreciated.

Quote:

Originally Posted by ECO

This is cake in R...

Depending on how large the datasets are, read them into R. The example below (using read.table) is just for using the clipboard in MacOSX (highlight in excel, copy, then run that line). The equiv in windows is read.table("clipboard",sep='\t',header=TRUE).

Code:

> df1 <- read.table(pipe("pbpaste"),header=TRUE)
> df1
  Gene Change    Padj
1    1    212 0.00060
2    2     99 0.09000
3    3     15 0.70000
4    4     59 0.45600
5    5    418 0.00001
6    6    566 0.00000
> df2 <- read.table(pipe("pbpaste"),header=TRUE)
> df2
  Gene Statistic    FDR
1    1        28 0.0005
2    2        17 0.0070
3    3         3 0.8100
4    4         8 0.2800
5    5        48 0.0020
6    6       111 0.0000
> df3 <- merge(df1,df2)
> df3
  Gene Change    Padj Statistic    FDR
1    1    212 0.00060        28 0.0005
2    2     99 0.09000        17 0.0070
3    3     15 0.70000         3 0.8100
4    4     59 0.45600         8 0.2800
5    5    418 0.00001        48 0.0020
6    6    566 0.00000       111 0.0000
> df3[df3$Padj< 0.05 & df3$FDR < 0.05,]
  Gene Change  Padj Statistic   FDR
1    1    212 6e-04        28 5e-04
5    5    418 1e-05        48 2e-03
6    6    566 0e+00       111 0e+00
> write.csv(df3[df3$Padj< 0.05 & df3$FDR < 0.05,])
"","Gene","Change","Padj","Statistic","FDR"
"1",1,212,6e-04,28,5e-04
"5",5,418,1e-05,48,0.002
"6",6,566,0,111,0

ECO · 10-03-2011, 02:06 PM

Just need to tell merge what to do...(note I added two extra genes into df1)

Code:

> df1
  Gene Change    Padj
1    1    212 0.00060
2    2     99 0.09000
3    3     15 0.70000
4    4     59 0.45600
5    5    418 0.00001
6    6    566 0.00000
7    7    123 0.00000
8    8    125 0.04000
> df2
  Gene Statistic    FDR
1    1        28 0.0005
2    2        17 0.0070
3    3         3 0.8100
4    4         8 0.2800
5    5        48 0.0020
6    6       111 0.0000
> merge(df1,df2,on=Gene)
  Gene Statistic    FDR Change    Padj
1    1        28 0.0005    212 0.00060
2    2        17 0.0070     99 0.09000
3    3         3 0.8100     15 0.70000
4    4         8 0.2800     59 0.45600
5    5        48 0.0020    418 0.00001
6    6       111 0.0000    566 0.00000

edit: Looks like "on" isn't necessary...something else going on with your data? Merge is pretty smart by default...

byou678 · 10-04-2011, 08:42 PM

Thanks again ECO. Yes, I like using R too and I have figured it out.

Quote:

Originally Posted by ECO

Just need to tell merge what to do...(note I added two extra genes into df1)

Code:

> df1
  Gene Change    Padj
1    1    212 0.00060
2    2     99 0.09000
3    3     15 0.70000
4    4     59 0.45600
5    5    418 0.00001
6    6    566 0.00000
7    7    123 0.00000
8    8    125 0.04000
> df2
  Gene Statistic    FDR
1    1        28 0.0005
2    2        17 0.0070
3    3         3 0.8100
4    4         8 0.2800
5    5        48 0.0020
6    6       111 0.0000
> merge(df1,df2,on=Gene)
  Gene Statistic    FDR Change    Padj
1    1        28 0.0005    212 0.00060
2    2        17 0.0070     99 0.09000
3    3         3 0.8100     15 0.70000
4    4         8 0.2800     59 0.45600
5    5        48 0.0020    418 0.00001
6    6       111 0.0000    566 0.00000

edit: Looks like "on" isn't necessary...something else going on with your data? Merge is pretty smart by default...

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Bug? duplicated genes in cufflinks output genes.expr	silin284	Bioinformatics	3	05-18-2014 12:19 AM
Common elements within two set of miRNAs	Giorgio C	Bioinformatics	1	01-05-2012 10:35 AM
Fastq sorting and common part?	stoker	Bioinformatics	0	07-07-2011 01:53 AM
very common mutation not in SNP databases	fpepin	Bioinformatics	17	02-11-2011 06:14 PM
edgeR with no replication (Common disp or poisson)	sergio	Bioinformatics	3	05-24-2010 09:53 PM

10-03-2011, 08:53 AM	#1
byou678 Member Location: Maryland Join Date: Aug 2011 Posts: 52	How to extract Common Genes from 2 spreadsheets Now I have two Excel spreadsheets, parts of them are as below: Spreadsheet1 Gene Change Padj 1 212 0.0006 2 99 0.09 3 15 0.7 4 59 0.456 5 418 0.00001 6 566 0 Spreadsheet2 Gene Statistic FDR 1 28 0.0005 2 17 0.007 3 3 0.81 4 8 0.28 5 48 0.002 6 111 0 Using Excel, how can I create one new spreadsheet that only includes the Genes with both Padj and FDR are less than 0.05? And how can I do this in any other statsitical software, Like R, SPSS? Thanks very much for any idea for any question!