FOR loop with statistical analysis for microarray data

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20111027/6157d5c5/attachment.pl>
In your loop, you assign, for example, pv twice
pv=w$p.value # pv is scalar
pv[i]= w[2]   # pv is a vector

give an example for the point

pv=1
pv[5]=2
pv
[1]  1 NA NA NA  2

This may not be what you want.

Weidong
y1,y2 were designed as follow,from the original data file :
y1=data[,1:2]
y2=data[,3:4]

...I am a bit confused with what "redesign the whole vector again" and
"specify indices twice " actually mean?...could u point it out in the
script?

Thanks so much I really appreciate it!!

Sent from my -DROID-

On Oct 26, 2011 6:56 PM, "Weidong Gu" ?wrote:
If you provide an example data (y1 and y2 in the loop), you might have
got specific helps already. A few things in your loop seem suspicious.
fc and pv are vectors, and in each loop you redesigned the whole
vectors and specific indices twice. That may cause your problems.

Weidong Gu

On Wed, Oct 26, 2011 at 4:56 PM, Seb <seba.bat at gmail.com> wrote:
hi all

i started recently using R and i found myself stuck when i try to
analyze microarray data.

i use the "affy" package to obtain ?the intensities of the probes, i
have two CTRs and two treated.

?HG.U133A.Experiment1.CEL HG.U133A.Experiment2.CEL
HG.U133A_Control1.CEL HG.U133A_Control2.CEL
1007_s_at ? ? ? ? ? ? ? 2156.23115 ? ? ? ? ? ? ? ?467.75615
?364.60615 ? ? ? ? ? ? 362.11865
1053_at ? ? ? ? ? ? ? ? ? 88.76368 ? ? ? ? ? ? ? ? 93.58436
?438.49365 ? ? ? ? ? ? 357.75615
117_at ? ? ? ? ? ? ? ? ? 144.00743 ? ? ? ? ? ? ? ?101.26120
?95.11117 ? ? ? ? ? ? 107.01623
121_at ? ? ? ? ? ? ? ? ? 551.36865 ? ? ? ? ? ? ? ?639.45615
?456.66865 ? ? ? ? ? ? 435.95615
1255_g_at ? ? ? ? ? ? ? ? 65.33164 ? ? ? ? ? ? ? ? 18.39570
?14.22565 ? ? ? ? ? ? ?20.74632
1294_at ? ? ? ? ? ? ? ? ?106.19083 ? ? ? ? ? ? ? ?169.69369
?78.15722 ? ? ? ? ? ? ?81.14689

i divided the first two columns in two data.frames to divide Experim and
CTRs
then, i created a FOR loop to create a vector per each row containing
a vector with two values per each gene and i wanted to do a
Wilcox.test to obtain the significant genes..BUT i get a list of NULL
like you can see here
..the first row works but then i get NULL down till the end of the
array...
? ? ? ? ? ? ? ?fc ? ? ? ?pv
[1,] "1007_s_at" -20.248 ? 0.4664612
[2,] "1053_at" ? -344.7132 NULL
[3,] "117_at" ? ?NULL ? ? ?NULL
[4,] "121_at" ? ?NULL ? ? ?NULL
[5,] "1255_g_at" NULL ? ? ?NULL
[6,] "1294_at" ? NULL ? ? ?NULL

the script i used is:
===================
fc=0
pv=0
for (i in 1:nrow(data))
{
? ? ? ?v1= c(y1[i,1], y1[i,2])
? ? ? ?v2= c(y2[i,1], y2[1,2])
? ? ? ?fc=v1-v2
? ? ? ?w=t.test(v1,v2)
? ? ? ?pv=w$p.value
? ? ? ?fc[i]= w[1]
? ? ? ?pv[i]= w[2]
}

results = cbind(row.names(y1), fc, pv)

head(results)

================

what did i do wrong? i can't find a way around this!!!

thanks so much!!!

Seb

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

? ? ? ?[[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

thanks for the replies.... so, i remove the indices pv[i] and
fc[i]..but when i run it like this:
===============
fc=0
pv=0
for (i in 1:nrow(data))
{

	v1= c(y1[i,1], y1[i,2])	
	v2= c(y2[i,1], y2[1,2])
	fc=v1-v2
	w=t.test(v1,v2)
	pv=w$p.value
}

results = cbind(row.names(y1), fc, pv)

head(results)
=================

i get

===========
Warning message:
In cbind(row.names(y1), fc, pv) :
  number of rows of result is not a multiple of vector length (arg 2)
=============

and the "fc" values are repeated over until the end of the rows and
the "pv" is the same all across the samples...how can i fix it?!

...i apologize if it's a silly situation but i'm new to this and can't
get my head around it!

thanks so much!!
In your loop, you assign, for example, pv twice
pv=w$p.value # pv is scalar
pv[i]= w[2] ? # pv is a vector

give an example for the point

pv=1
pv[5]=2
pv
[1] ?1 NA NA NA ?2

This may not be what you want.

Weidong

On Thu, Oct 27, 2011 at 7:40 AM, Seb <seba.bat at gmail.com> wrote:
y1,y2 were designed as follow,from the original data file :
y1=data[,1:2]
y2=data[,3:4]

...I am a bit confused with what "redesign the whole vector again" and
"specify indices twice " actually mean?...could u point it out in the
script?

Thanks so much I really appreciate it!!

Sent from my -DROID-

On Oct 26, 2011 6:56 PM, "Weidong Gu" ?wrote:
If you provide an example data (y1 and y2 in the loop), you might have
got specific helps already. A few things in your loop seem suspicious.
fc and pv are vectors, and in each loop you redesigned the whole
vectors and specific indices twice. That may cause your problems.

Weidong Gu

On Wed, Oct 26, 2011 at 4:56 PM, Seb <seba.bat at gmail.com> wrote:
hi all

i started recently using R and i found myself stuck when i try to
analyze microarray data.

i use the "affy" package to obtain ?the intensities of the probes, i
have two CTRs and two treated.

?HG.U133A.Experiment1.CEL HG.U133A.Experiment2.CEL
HG.U133A_Control1.CEL HG.U133A_Control2.CEL
1007_s_at ? ? ? ? ? ? ? 2156.23115 ? ? ? ? ? ? ? ?467.75615
?364.60615 ? ? ? ? ? ? 362.11865
1053_at ? ? ? ? ? ? ? ? ? 88.76368 ? ? ? ? ? ? ? ? 93.58436
?438.49365 ? ? ? ? ? ? 357.75615
117_at ? ? ? ? ? ? ? ? ? 144.00743 ? ? ? ? ? ? ? ?101.26120
?95.11117 ? ? ? ? ? ? 107.01623
121_at ? ? ? ? ? ? ? ? ? 551.36865 ? ? ? ? ? ? ? ?639.45615
?456.66865 ? ? ? ? ? ? 435.95615
1255_g_at ? ? ? ? ? ? ? ? 65.33164 ? ? ? ? ? ? ? ? 18.39570
?14.22565 ? ? ? ? ? ? ?20.74632
1294_at ? ? ? ? ? ? ? ? ?106.19083 ? ? ? ? ? ? ? ?169.69369
?78.15722 ? ? ? ? ? ? ?81.14689

i divided the first two columns in two data.frames to divide Experim and
CTRs
then, i created a FOR loop to create a vector per each row containing
a vector with two values per each gene and i wanted to do a
Wilcox.test to obtain the significant genes..BUT i get a list of NULL
like you can see here
..the first row works but then i get NULL down till the end of the
array...
? ? ? ? ? ? ? ?fc ? ? ? ?pv
[1,] "1007_s_at" -20.248 ? 0.4664612
[2,] "1053_at" ? -344.7132 NULL
[3,] "117_at" ? ?NULL ? ? ?NULL
[4,] "121_at" ? ?NULL ? ? ?NULL
[5,] "1255_g_at" NULL ? ? ?NULL
[6,] "1294_at" ? NULL ? ? ?NULL

the script i used is:
===================
fc=0
pv=0
for (i in 1:nrow(data))
{
? ? ? ?v1= c(y1[i,1], y1[i,2])
? ? ? ?v2= c(y2[i,1], y2[1,2])
? ? ? ?fc=v1-v2
? ? ? ?w=t.test(v1,v2)
? ? ? ?pv=w$p.value
? ? ? ?fc[i]= w[1]
? ? ? ?pv[i]= w[2]
}

results = cbind(row.names(y1), fc, pv)

head(results)

================

what did i do wrong? i can't find a way around this!!!

thanks so much!!!

Seb

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

? ? ? ?[[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Looks like you have problems with indexing. In your code, fc is a
vector of length 2, pv is a scalar(it only keeps the last value of the
loop), then you try to cbind different lengths of vectors
(row.names(y1), fc and pv). cbind only works for vectors with the same
length.

I don't know what you want to achieve, but you can keep p.value using
pv[i]. In the previous code, you reassiged pv before pv[i]. So your
1st assignment of pv wiped out pv vector. You need to read helps about
indexing, vectors.

Weidong
thanks for the replies.... so, i remove the indices pv[i] and
fc[i]..but when i run it like this:
===============
fc=0
pv=0
for (i in 1:nrow(data))
{

? ? ? ?v1= c(y1[i,1], y1[i,2])
? ? ? ?v2= c(y2[i,1], y2[1,2])
? ? ? ?fc=v1-v2
? ? ? ?w=t.test(v1,v2)
? ? ? ?pv=w$p.value
}

results = cbind(row.names(y1), fc, pv)

head(results)
=================

i get

===========
Warning message:
In cbind(row.names(y1), fc, pv) :
?number of rows of result is not a multiple of vector length (arg 2)
=============

and the "fc" values are repeated over until the end of the rows and
the "pv" is the same all across the samples...how can i fix it?!

...i apologize if it's a silly situation but i'm new to this and can't
get my head around it!

thanks so much!!

On Thu, Oct 27, 2011 at 9:02 AM, Weidong Gu <anopheles123 at gmail.com> wrote:
In your loop, you assign, for example, pv twice
pv=w$p.value # pv is scalar
pv[i]= w[2] ? # pv is a vector

give an example for the point

pv=1
pv[5]=2
pv
[1] ?1 NA NA NA ?2

This may not be what you want.

Weidong

On Thu, Oct 27, 2011 at 7:40 AM, Seb <seba.bat at gmail.com> wrote:
y1,y2 were designed as follow,from the original data file :
y1=data[,1:2]
y2=data[,3:4]

...I am a bit confused with what "redesign the whole vector again" and
"specify indices twice " actually mean?...could u point it out in the
script?

Thanks so much I really appreciate it!!

Sent from my -DROID-

On Oct 26, 2011 6:56 PM, "Weidong Gu" ?wrote:
If you provide an example data (y1 and y2 in the loop), you might have
got specific helps already. A few things in your loop seem suspicious.
fc and pv are vectors, and in each loop you redesigned the whole
vectors and specific indices twice. That may cause your problems.

Weidong Gu

On Wed, Oct 26, 2011 at 4:56 PM, Seb <seba.bat at gmail.com> wrote:
hi all

i started recently using R and i found myself stuck when i try to
analyze microarray data.

i use the "affy" package to obtain ?the intensities of the probes, i
have two CTRs and two treated.

?HG.U133A.Experiment1.CEL HG.U133A.Experiment2.CEL
HG.U133A_Control1.CEL HG.U133A_Control2.CEL
1007_s_at ? ? ? ? ? ? ? 2156.23115 ? ? ? ? ? ? ? ?467.75615
?364.60615 ? ? ? ? ? ? 362.11865
1053_at ? ? ? ? ? ? ? ? ? 88.76368 ? ? ? ? ? ? ? ? 93.58436
?438.49365 ? ? ? ? ? ? 357.75615
117_at ? ? ? ? ? ? ? ? ? 144.00743 ? ? ? ? ? ? ? ?101.26120
?95.11117 ? ? ? ? ? ? 107.01623
121_at ? ? ? ? ? ? ? ? ? 551.36865 ? ? ? ? ? ? ? ?639.45615
?456.66865 ? ? ? ? ? ? 435.95615
1255_g_at ? ? ? ? ? ? ? ? 65.33164 ? ? ? ? ? ? ? ? 18.39570
?14.22565 ? ? ? ? ? ? ?20.74632
1294_at ? ? ? ? ? ? ? ? ?106.19083 ? ? ? ? ? ? ? ?169.69369
?78.15722 ? ? ? ? ? ? ?81.14689

i divided the first two columns in two data.frames to divide Experim and
CTRs
then, i created a FOR loop to create a vector per each row containing
a vector with two values per each gene and i wanted to do a
Wilcox.test to obtain the significant genes..BUT i get a list of NULL
like you can see here
..the first row works but then i get NULL down till the end of the
array...
? ? ? ? ? ? ? ?fc ? ? ? ?pv
[1,] "1007_s_at" -20.248 ? 0.4664612
[2,] "1053_at" ? -344.7132 NULL
[3,] "117_at" ? ?NULL ? ? ?NULL
[4,] "121_at" ? ?NULL ? ? ?NULL
[5,] "1255_g_at" NULL ? ? ?NULL
[6,] "1294_at" ? NULL ? ? ?NULL

the script i used is:
===================
fc=0
pv=0
for (i in 1:nrow(data))
{
? ? ? ?v1= c(y1[i,1], y1[i,2])
? ? ? ?v2= c(y2[i,1], y2[1,2])
? ? ? ?fc=v1-v2
? ? ? ?w=t.test(v1,v2)
? ? ? ?pv=w$p.value
? ? ? ?fc[i]= w[1]
? ? ? ?pv[i]= w[2]
}

results = cbind(row.names(y1), fc, pv)

head(results)

================

what did i do wrong? i can't find a way around this!!!

thanks so much!!!

Seb

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

? ? ? ?[[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.