Combine 100K or 500K sub-arrays
12 views
Skip to first unread message
Cheng Li
Dec 30, 2008, 7:22:37 AM12/30/08
to dChip Software
In dc...@yahoogroups.com, "Sam Volchenboum" <accounts@...> wrote:
I've been playing with this for a few hours, and I'd like to make
sure I'm doing this correctly. I'm trying to combine the 50K HIND and
50K XBA sub-arrays.
I have 9 samples. Each had a blood, tumor and cell-line. For each,
there are XBA and HIND arrays. So, there are 27 arrays for each
enzyme.
1. I analyzed the data of each sub-array separately through
normalization and model-based signal to obtain DCP files
2. Created a new folder and copied all the DCPs into it. I renamed
all the HIND files to have X's but with a _2 on the end. For
instance, if the original XBA file was 08X_1355T_x.DCP, then the
corresponding HIND file was: 08H_1355T_h.DCP, which I renamed to:
08X_1355T_x_2.DCP.
3. Analysis/Open Group
Data Files:
Group name: XBA
File Type: DCP
Nothing else checked
Other Informtaion:
CDF File:
CDF File: Mapping50K_Xba240.CDF
Subarray CDF: Mapping50K_Hind240.CDF
Array Type: SNP 100K
Information Files:
Gene or SNP: Mapping100K_Xba snp info ** IS THIS RIGHT?
Should it be the
combined info?
Sample: XBA_Sample_Info ** contains the XBA samples
and none of the _2 samples.
Should I create a new Sample Info
file that contains all 54
(27 XBA and 27 HIND)?
Options: Uncheck "Load Probe Data in Memory"
Change Working Directory
4. Tools/Array List File
Load XBA array list ** IS THIS RIGHT?
There are only the 27 XBA arrays.
Should I make a new array list file with
both the XBA and HIND arrays?
Of course, the HIND arrays are just XBA_2.
5. Analysis/Chromosome
Genome Information File: "100k genome info AfAm june 05 hg17"
** USE THIS ONE INSTEAD OF THE
50K XBA, right?
Reference gene file: "hu refGene hg17"
Cytoband file: "cytoBand hg17 sorted"
Analysis method: "Copy number & LOH"
The analysis seems to complete (no errors). But the LOH and CN data
do not look right - there are whole chromosomes with LOH, for
instance. Also, when I try to go to Chromosome/Summary Plot, the
program crashes every time. This does not happen with I not try to
combine arrays.
I'm wondering if I need to create new Sample Info and Array List
files with all 54 arrays (the XBA and the HIND (now XBA_2)).
Any advice or changes are appreciated. Let me know if it would be
helpful for me to post my log files.
Thanks.
sam
From: David Twomey
Sent: Tuesday, May 22, 2007 7:07 AM
Subject: [dChip] Combining the NSP & STY sets for the 500k array
Cheng,
I saw this email on the dChip mailing list. GenePattern has modules
that can do what she needs. We have 2 modules, "MergeRows" which will
merge the NSP and STY together and "SNPFileSorter" which will resort
the
file by Chromosome and location. If you don't yet have this
functionality in dChipSNP then this is a possible solution to her
needs
in the meantime.
Best Regards
David
Subject: [dChip] Combining the NSP & STY sets for the 500k array
Date: Mon, 21 May 2007 11:52:33 -0700
From: Corina Shtir
Dear all,
Can you please help me figure out how to combine the NSP & STY chip
data for the 500k Affy array type? I normalized the separate NSP and
STY files and then I tried exporting the expression values for the
filtered genes of each set. Since our data has 303 people, every time
I try opening the .XLS file with the 'filtered genes expression'
data,
the file does not load completely because it surpasses the nr. of
columns allowed by Excel.
How can I get around this?
Many thanks.
Corina
RE: [dChip] Combine subarrays - error
Chris,
Probably you need to open Sty files in a group and also do “Analysis/
Normalize & Model”.
Cheng
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of christopher....@gsk.com
Sent: Tuesday, May 08, 2007 9:55 AM
Subject: [dChip] Combine subarrays - error
Hi Cheng,
I am working with 500K SNP chip data and I am interested in obtaining
LOH calls. I am having a little difficulty with combining the sub-
arrays as outlined under
http://biosun1.harvard.edu/complab/dchip/combine%20chip.htm#sub_chip
I have obtained .dcp files for both nsp and sty data separately, with
the dcp files for the nsp and sty files being named FILE.dcp and
FILE_2.dcp, respectively. When I go to open the combined group (as
outlined in the manual), the CDF files are opened correctly as
expected below:
{Open group combine
Found this array type as SNP array
Read in existing binary CDF file C:\Documents and Settings
\cqm64023\Desktop\dchip_pilot\Mapping250K_Nsp.cdf.bin (file format
4)...
Reading the CDF file of the subarray...
Found this array type as SNP array
Read in existing binary CDF file C:\Documents and Settings
\cqm64023\Desktop\dchip_pilot\Mapping250K_Sty.cdf.bin (file format
4)...
And the nsp files are opened as expected:
Found C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\BC1.dcp
Accessing 'C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\BC1.dcp' (file format 4)
Found C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\CAPAN2.dcp
Accessing 'C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\CAPAN2.dcp' (file format 4)
but when the sty files are to be opened (with the convention
FILE_2.dcp) I get this error:
Load the DCP files of the subarray...
Accessing 'C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\BC1_2.dcp' (file format 4)
This DCP file is not normalized or modelled. First open this subarray
type in a separate group and do normalization and modelling.
Opening group not completed}
I am a little confused because I did generate the dcp files for the
subarray (sty files) separately. Any ideas on how I can resolve this
and combine my data?
Chris
Probably the combined data file is not saved in tab-delimited text
format. -- Cheng
Wed Mar 14, 2007 8:00 pm
--- In dc...@yahoogroups.com, qiu_wen <no_reply@...> wrote:
Dear all,
I'm trying to combine 250K Nsp and StyI SNP array data in dCHIP.
However, I kept getting error messgage "MFC application has
encountered a problem" and dCHIP window was forced to close. I
thought
it could be a memory problem and tried to follow the instructions on
the dCHIP web. But this problem keeps coming up. I'm not sure whether
it's a computer problem or dCHIP problem.
Cheers,
Wen
I think it makes sense but I don’t know how much it can help, since
the Xba and Hind array of the same sample may have different
brightness as well. You may try this to see if it improves.
Cheng Li
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of Wei, CHEN
Sent: Tuesday, March 28, 2006 5:15 AM
To: dc...@yahoogroups.com
Subject: [dChip] Combining 100k Xba and Hind Arrays
Hi, Dear Cheng Li:
I have some questions about the combining procedure for 100k SNP
Mapping arrays.
As I read from manual, the combination should be done after
Normalization and MBEI for each Chip types group.
Here is my question:
Would it be better to use the same sample as baseline for both Xba
and
Hind chips normalization? Would it cause bias for combination if the
chip intensities of baselines are different?
Thanks,
Chen Wei
Hi Jian,
In dchip we can only normalize and compute signals before combing the
two arrays. Since the two sub-arrays have different probes and will
not be modeled together for computing signal, it’s not essential to
normalize them at the same time.
dChip doesn’t read or store standard deviations from CEL files so
cannot export them.
Cheng Li
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of Wang, Jian
Sent: Tuesday, April 04, 2006 11:04 AM
To: dc...@yahoogroups.com
Subject: RE: [dChip] Combining 100k Xba and Hind Arrays
Dear Cheng,
I am doing copy number analysis using dCHIP for 100K SNP array data.
I've got a question about the combined analysis following a previous
discussion (attached at the end).
It seems that the "Combine sub-arrays for analysis" procedures
described in dChip manual can be done only after intensities from the
Xba array and intensities from the Hind array have been normalized
independently (within each sub-array). I am just wondering if it is
possible to normalize the intensity signals after these two arrays
are
combined together. It does not seem to be possible to do this since
the SNPs in the 50K Hind array and in the 50K Xba array are totally
different and normalization of SNP intensities is based on all
samples' intensities for each SNP. Please let me know if my
understanding of this normalization process is correct.
I also have an additional question about the standard deviation of
probe intensities. dCHIP can output the unnormalized and normalized
median intensity for each array. I am just wonderinf if dCHIP can
also
output the standard deviation of probe intensities for each array.
Thanks,
Jian
Thu Apr 6, 2006 8:21 pm
Hi Jason,
This is caused by having header information in the exteranl data
file.
You can either delete header (in the beginning and in the middle for
the 2nd subchip) or uncheck 'Tools/Export expression data/Include
header information' to re-export the file from dChip.
Thanks for pointing out the issue.
Cheng Li
From: jcorneveaux@tgen
Sent: Thursday, April 06, 2006 7:02 PM
Subject: Combine SNP arrays 500K
Dear Cheng,
I am trying to combine the Mapping250K Sty and Nsp arrays as outlined
on the dChip website. Each time I “Get external data” the program
crashes after reading in the expression data file. I have attached
the
error file if this may help. I also saw on the website that combing
arrays on the 100K and 500K is difficult, does that mean it is
impossible? Perhaps this is the issue? I have tried this on several
computers with 4GB of memory and dual Xeon processors so I don’t
think
it is a memory/hardware issue. Any help would be greatly appreciated.
Bests,
Jason Corneveaux
Wed Aug 23, 2006 6:47 pm
Hi Wei,
I think the detection resolution will increase after you combine.
However the two sub arrays may not give similar estimate of copy
change in the same region, and may increase the noise level a little.
You can compare and decide whether to combine.
Cheng
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of wwresearch2006
Sent: Tuesday, August 22, 2006 4:14 PM
To: dc...@yahoogroups.com
Subject: [dChip] combine Affy NSP and STY 500K for copy number/LOH
analysis?
Hi, Dr. Li:
I have a question regarding copy number/LOH analysis: if I combine
the
500K NSP and STY, do I gain any accuracy in copy number/LOH analysis?
Especially with HMM, since there will be more SNPs/more information
available?
thanks
wei
Fri Jan 5, 2007 12:50 pm
RE: [dChip] Pb : combine subarray 500k (NSP/STY)
Fjulienf,
Yes you need all subarray pairs of each sample. In this case you can
move the two unpaired NSP arrays out of the directory before
combining.
Cheng
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of fjulienf
Sent: Friday, January 05, 2007 4:15 AM
Subject: [dChip] Pb : combine subarray 500k (NSP/STY)
Dear all
I have 36 patients samples (19 NSP and 17 STY) files (2 STY patients
samples are BAD) and i am interested to combine this differents
samples for my study.
I try to combine the sample as describe in
http://www.biostat.harvard.edu/complab/dchip/combine%20chip.htm#sub_chip
but an error occurs :
{Open group ONE
Found this array type as SNP array
Read in existing binary CDF file
C:\julien\500K\Mapping250K_Nsp.cdf.bin (file format 4)...
Reading the CDF file of the subarray...
Found this array type as SNP array
Read in existing binary CDF file
C:\julien\500K\Mapping250K_Sty.cdf.bin (file format 4)...
Search and extract PM/MM data from 'DCP' files of chip type
'Mapping250K_Nsp' in
C:\julien\500K\OLD...
Found C:\julien\500K\OLD\NVQ.dcp
Accessing 'C:\julien\500K\OLD\NVQ.dcp' (file format 4)
.
//
.
Found C:\julien\500K\OLD\OIE.dcp
Accessing 'C:\julien\500K\OLD\OIE.dcp' (file format 4)
Load the DCP files of the subarray...
Reading file 'C:\julien\500K\OLD\NVQ_2.dcp' failed. Check if the
file exists or match the name of the first sub array
Opening group not completed}
In fact I do not have the sample STY for NVQ (BAD sample) "NVQ_2.dcp"
so ...
What can I do ? Do I need pairs for all samples (NSP and STY)?
Thanks
Wed Sep 13, 2006 2:34 pm
Jason,
I think so, if you have matching 100K and 500K files for the same
samples.
You will need to go through external file to row-wise combined data
exported
by "Tools/Export expression value", and then read them back by "Get
external
data". Similarly row-wise combine genome info files.
Cheng
From: jcorneveaux@tgen
Sent: Wednesday, September 13, 2006 1:33 PM
To: c...@hsph.harvard.edu
Subject: 600K dChip?
Hi Cheng,
I have a quick question for you - Do you think I could combine 100K
and 500K
data in dChip? I think the process would be similar to combining two
arrays
(like the xba and hindi). I would have to create a new genome
information
file and then combine the 100K and 500K data. I can see how this
might
work,
but I wonder if it will work, so I thought I'd ask you...
Thanks Cheng,
Jason
Fri May 26, 2006 6:40 pm
--- In dc...@yahoogroups.com, "arrayprofile" <arrayprofile@...>
wrote:
Hi Cheng,
Thanks for the previous help, now dChip is reading in all samples
without any problem.
since I am using 100k SNP arrays, so I definitely want to combine the
2
sub-arrays. In the following paragraph about combining subarrays, it
seems that the 2 DCP files should have the same name, except ,
except "_2" at the end of the 2nd sub-arrays. However, my case is
that
each sub-arrays are discriminated by either "_Xba_" or "_Hind_" in
the
names. Can I still follow your instructions to combine the 2 sub-
arrays
without using external data files?
Thu Mar 2, 2006 11:12 am
I agree. This could happen when the array list file used for
exporting
the two subarrays are not the same or aligned.
Cheng Li
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of Natalie Twine
Sent: Thursday, March 02, 2006 6:19 AM
To: dc...@yahoogroups.com
Subject: Re: [dChip] Crash with combined 100k
hi alfredo,
open the combined expression file in notepad/excel after you have
created it, before you use it in 'get external data'. check that the
columns are all in order and all the data has been exported correctly
-
I have had this error message a few times - and it was always due to
something being incorrect in the combined expression file. I imagine
as
it crashes at 59, 000, the second 50K array output data you are
exporting to the combined expressiong file is not in the same format
as
the first 50K array output data.
natalie
ALFREDO HIDALGO wrote:
I am trying to open a combined file of the 100k set, after appending
one expresion info file to the other, following the instructions, but
when the program reaches line 59,000 it crashes and closes.
Sny clue about why this is happening?
Thanks!!
Alfredo Hidalgo-Miranda
Instituto Nacional de Medicina Genómica
Sun Mar 5, 2006 11:14 pm
--- In dc...@yahoogroups.com, ALFREDO HIDALGO <alfhm@...> wrote:
Hi!!
Thank you all for your help, the problem was, as you
suggested, that I had some chips that were not paired,
I have solved this and it runs fine.
Thanks again!!
--- mei xu <mmmxxxuuu2003@...> wrote:
You can put all the files into a database, then
query whatever you want. If you are not familiar
with database and need help, please send your
questions.
I've been playing with this for a few hours, and I'd like to make
sure I'm doing this correctly. I'm trying to combine the 50K HIND and
50K XBA sub-arrays.
I have 9 samples. Each had a blood, tumor and cell-line. For each,
there are XBA and HIND arrays. So, there are 27 arrays for each
enzyme.
1. I analyzed the data of each sub-array separately through
normalization and model-based signal to obtain DCP files
2. Created a new folder and copied all the DCPs into it. I renamed
all the HIND files to have X's but with a _2 on the end. For
instance, if the original XBA file was 08X_1355T_x.DCP, then the
corresponding HIND file was: 08H_1355T_h.DCP, which I renamed to:
08X_1355T_x_2.DCP.
3. Analysis/Open Group
Data Files:
Group name: XBA
File Type: DCP
Nothing else checked
Other Informtaion:
CDF File:
CDF File: Mapping50K_Xba240.CDF
Subarray CDF: Mapping50K_Hind240.CDF
Array Type: SNP 100K
Information Files:
Gene or SNP: Mapping100K_Xba snp info ** IS THIS RIGHT?
Should it be the
combined info?
Sample: XBA_Sample_Info ** contains the XBA samples
and none of the _2 samples.
Should I create a new Sample Info
file that contains all 54
(27 XBA and 27 HIND)?
Options: Uncheck "Load Probe Data in Memory"
Change Working Directory
4. Tools/Array List File
Load XBA array list ** IS THIS RIGHT?
There are only the 27 XBA arrays.
Should I make a new array list file with
both the XBA and HIND arrays?
Of course, the HIND arrays are just XBA_2.
5. Analysis/Chromosome
Genome Information File: "100k genome info AfAm june 05 hg17"
** USE THIS ONE INSTEAD OF THE
50K XBA, right?
Reference gene file: "hu refGene hg17"
Cytoband file: "cytoBand hg17 sorted"
Analysis method: "Copy number & LOH"
The analysis seems to complete (no errors). But the LOH and CN data
do not look right - there are whole chromosomes with LOH, for
instance. Also, when I try to go to Chromosome/Summary Plot, the
program crashes every time. This does not happen with I not try to
combine arrays.
I'm wondering if I need to create new Sample Info and Array List
files with all 54 arrays (the XBA and the HIND (now XBA_2)).
Any advice or changes are appreciated. Let me know if it would be
helpful for me to post my log files.
Thanks.
sam
From: David Twomey
Sent: Tuesday, May 22, 2007 7:07 AM
Subject: [dChip] Combining the NSP & STY sets for the 500k array
Cheng,
I saw this email on the dChip mailing list. GenePattern has modules
that can do what she needs. We have 2 modules, "MergeRows" which will
merge the NSP and STY together and "SNPFileSorter" which will resort
the
file by Chromosome and location. If you don't yet have this
functionality in dChipSNP then this is a possible solution to her
needs
in the meantime.
Best Regards
David
Subject: [dChip] Combining the NSP & STY sets for the 500k array
Date: Mon, 21 May 2007 11:52:33 -0700
From: Corina Shtir
Dear all,
Can you please help me figure out how to combine the NSP & STY chip
data for the 500k Affy array type? I normalized the separate NSP and
STY files and then I tried exporting the expression values for the
filtered genes of each set. Since our data has 303 people, every time
I try opening the .XLS file with the 'filtered genes expression'
data,
the file does not load completely because it surpasses the nr. of
columns allowed by Excel.
How can I get around this?
Many thanks.
Corina
RE: [dChip] Combine subarrays - error
Chris,
Probably you need to open Sty files in a group and also do “Analysis/
Normalize & Model”.
Cheng
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of christopher....@gsk.com
Sent: Tuesday, May 08, 2007 9:55 AM
Subject: [dChip] Combine subarrays - error
Hi Cheng,
I am working with 500K SNP chip data and I am interested in obtaining
LOH calls. I am having a little difficulty with combining the sub-
arrays as outlined under
http://biosun1.harvard.edu/complab/dchip/combine%20chip.htm#sub_chip
I have obtained .dcp files for both nsp and sty data separately, with
the dcp files for the nsp and sty files being named FILE.dcp and
FILE_2.dcp, respectively. When I go to open the combined group (as
outlined in the manual), the CDF files are opened correctly as
expected below:
{Open group combine
Found this array type as SNP array
Read in existing binary CDF file C:\Documents and Settings
\cqm64023\Desktop\dchip_pilot\Mapping250K_Nsp.cdf.bin (file format
4)...
Reading the CDF file of the subarray...
Found this array type as SNP array
Read in existing binary CDF file C:\Documents and Settings
\cqm64023\Desktop\dchip_pilot\Mapping250K_Sty.cdf.bin (file format
4)...
And the nsp files are opened as expected:
Found C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\BC1.dcp
Accessing 'C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\BC1.dcp' (file format 4)
Found C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\CAPAN2.dcp
Accessing 'C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\CAPAN2.dcp' (file format 4)
but when the sty files are to be opened (with the convention
FILE_2.dcp) I get this error:
Load the DCP files of the subarray...
Accessing 'C:\Documents and Settings\cqm64023\Desktop\dchip_pilot\CEL
\combine\BC1_2.dcp' (file format 4)
This DCP file is not normalized or modelled. First open this subarray
type in a separate group and do normalization and modelling.
Opening group not completed}
I am a little confused because I did generate the dcp files for the
subarray (sty files) separately. Any ideas on how I can resolve this
and combine my data?
Chris
Probably the combined data file is not saved in tab-delimited text
format. -- Cheng
Wed Mar 14, 2007 8:00 pm
--- In dc...@yahoogroups.com, qiu_wen <no_reply@...> wrote:
Dear all,
I'm trying to combine 250K Nsp and StyI SNP array data in dCHIP.
However, I kept getting error messgage "MFC application has
encountered a problem" and dCHIP window was forced to close. I
thought
it could be a memory problem and tried to follow the instructions on
the dCHIP web. But this problem keeps coming up. I'm not sure whether
it's a computer problem or dCHIP problem.
Cheers,
Wen
I think it makes sense but I don’t know how much it can help, since
the Xba and Hind array of the same sample may have different
brightness as well. You may try this to see if it improves.
Cheng Li
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of Wei, CHEN
Sent: Tuesday, March 28, 2006 5:15 AM
To: dc...@yahoogroups.com
Subject: [dChip] Combining 100k Xba and Hind Arrays
Hi, Dear Cheng Li:
I have some questions about the combining procedure for 100k SNP
Mapping arrays.
As I read from manual, the combination should be done after
Normalization and MBEI for each Chip types group.
Here is my question:
Would it be better to use the same sample as baseline for both Xba
and
Hind chips normalization? Would it cause bias for combination if the
chip intensities of baselines are different?
Thanks,
Chen Wei
Hi Jian,
In dchip we can only normalize and compute signals before combing the
two arrays. Since the two sub-arrays have different probes and will
not be modeled together for computing signal, it’s not essential to
normalize them at the same time.
dChip doesn’t read or store standard deviations from CEL files so
cannot export them.
Cheng Li
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of Wang, Jian
Sent: Tuesday, April 04, 2006 11:04 AM
To: dc...@yahoogroups.com
Subject: RE: [dChip] Combining 100k Xba and Hind Arrays
Dear Cheng,
I am doing copy number analysis using dCHIP for 100K SNP array data.
I've got a question about the combined analysis following a previous
discussion (attached at the end).
It seems that the "Combine sub-arrays for analysis" procedures
described in dChip manual can be done only after intensities from the
Xba array and intensities from the Hind array have been normalized
independently (within each sub-array). I am just wondering if it is
possible to normalize the intensity signals after these two arrays
are
combined together. It does not seem to be possible to do this since
the SNPs in the 50K Hind array and in the 50K Xba array are totally
different and normalization of SNP intensities is based on all
samples' intensities for each SNP. Please let me know if my
understanding of this normalization process is correct.
I also have an additional question about the standard deviation of
probe intensities. dCHIP can output the unnormalized and normalized
median intensity for each array. I am just wonderinf if dCHIP can
also
output the standard deviation of probe intensities for each array.
Thanks,
Jian
Thu Apr 6, 2006 8:21 pm
Hi Jason,
This is caused by having header information in the exteranl data
file.
You can either delete header (in the beginning and in the middle for
the 2nd subchip) or uncheck 'Tools/Export expression data/Include
header information' to re-export the file from dChip.
Thanks for pointing out the issue.
Cheng Li
From: jcorneveaux@tgen
Sent: Thursday, April 06, 2006 7:02 PM
Subject: Combine SNP arrays 500K
Dear Cheng,
I am trying to combine the Mapping250K Sty and Nsp arrays as outlined
on the dChip website. Each time I “Get external data” the program
crashes after reading in the expression data file. I have attached
the
error file if this may help. I also saw on the website that combing
arrays on the 100K and 500K is difficult, does that mean it is
impossible? Perhaps this is the issue? I have tried this on several
computers with 4GB of memory and dual Xeon processors so I don’t
think
it is a memory/hardware issue. Any help would be greatly appreciated.
Bests,
Jason Corneveaux
Wed Aug 23, 2006 6:47 pm
Hi Wei,
I think the detection resolution will increase after you combine.
However the two sub arrays may not give similar estimate of copy
change in the same region, and may increase the noise level a little.
You can compare and decide whether to combine.
Cheng
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of wwresearch2006
Sent: Tuesday, August 22, 2006 4:14 PM
To: dc...@yahoogroups.com
Subject: [dChip] combine Affy NSP and STY 500K for copy number/LOH
analysis?
Hi, Dr. Li:
I have a question regarding copy number/LOH analysis: if I combine
the
500K NSP and STY, do I gain any accuracy in copy number/LOH analysis?
Especially with HMM, since there will be more SNPs/more information
available?
thanks
wei
Fri Jan 5, 2007 12:50 pm
RE: [dChip] Pb : combine subarray 500k (NSP/STY)
Fjulienf,
Yes you need all subarray pairs of each sample. In this case you can
move the two unpaired NSP arrays out of the directory before
combining.
Cheng
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of fjulienf
Sent: Friday, January 05, 2007 4:15 AM
Subject: [dChip] Pb : combine subarray 500k (NSP/STY)
Dear all
I have 36 patients samples (19 NSP and 17 STY) files (2 STY patients
samples are BAD) and i am interested to combine this differents
samples for my study.
I try to combine the sample as describe in
http://www.biostat.harvard.edu/complab/dchip/combine%20chip.htm#sub_chip
but an error occurs :
{Open group ONE
Found this array type as SNP array
Read in existing binary CDF file
C:\julien\500K\Mapping250K_Nsp.cdf.bin (file format 4)...
Reading the CDF file of the subarray...
Found this array type as SNP array
Read in existing binary CDF file
C:\julien\500K\Mapping250K_Sty.cdf.bin (file format 4)...
Search and extract PM/MM data from 'DCP' files of chip type
'Mapping250K_Nsp' in
C:\julien\500K\OLD...
Found C:\julien\500K\OLD\NVQ.dcp
Accessing 'C:\julien\500K\OLD\NVQ.dcp' (file format 4)
.
//
.
Found C:\julien\500K\OLD\OIE.dcp
Accessing 'C:\julien\500K\OLD\OIE.dcp' (file format 4)
Load the DCP files of the subarray...
Reading file 'C:\julien\500K\OLD\NVQ_2.dcp' failed. Check if the
file exists or match the name of the first sub array
Opening group not completed}
In fact I do not have the sample STY for NVQ (BAD sample) "NVQ_2.dcp"
so ...
What can I do ? Do I need pairs for all samples (NSP and STY)?
Thanks
Wed Sep 13, 2006 2:34 pm
Jason,
I think so, if you have matching 100K and 500K files for the same
samples.
You will need to go through external file to row-wise combined data
exported
by "Tools/Export expression value", and then read them back by "Get
external
data". Similarly row-wise combine genome info files.
Cheng
From: jcorneveaux@tgen
Sent: Wednesday, September 13, 2006 1:33 PM
To: c...@hsph.harvard.edu
Subject: 600K dChip?
Hi Cheng,
I have a quick question for you - Do you think I could combine 100K
and 500K
data in dChip? I think the process would be similar to combining two
arrays
(like the xba and hindi). I would have to create a new genome
information
file and then combine the 100K and 500K data. I can see how this
might
work,
but I wonder if it will work, so I thought I'd ask you...
Thanks Cheng,
Jason
Fri May 26, 2006 6:40 pm
--- In dc...@yahoogroups.com, "arrayprofile" <arrayprofile@...>
wrote:
Hi Cheng,
Thanks for the previous help, now dChip is reading in all samples
without any problem.
since I am using 100k SNP arrays, so I definitely want to combine the
2
sub-arrays. In the following paragraph about combining subarrays, it
seems that the 2 DCP files should have the same name, except ,
except "_2" at the end of the 2nd sub-arrays. However, my case is
that
each sub-arrays are discriminated by either "_Xba_" or "_Hind_" in
the
names. Can I still follow your instructions to combine the 2 sub-
arrays
without using external data files?
Thu Mar 2, 2006 11:12 am
I agree. This could happen when the array list file used for
exporting
the two subarrays are not the same or aligned.
Cheng Li
From: dc...@yahoogroups.com [mailto:dc...@yahoogroups.com] On Behalf
Of Natalie Twine
Sent: Thursday, March 02, 2006 6:19 AM
To: dc...@yahoogroups.com
Subject: Re: [dChip] Crash with combined 100k
hi alfredo,
open the combined expression file in notepad/excel after you have
created it, before you use it in 'get external data'. check that the
columns are all in order and all the data has been exported correctly
-
I have had this error message a few times - and it was always due to
something being incorrect in the combined expression file. I imagine
as
it crashes at 59, 000, the second 50K array output data you are
exporting to the combined expressiong file is not in the same format
as
the first 50K array output data.
natalie
ALFREDO HIDALGO wrote:
I am trying to open a combined file of the 100k set, after appending
one expresion info file to the other, following the instructions, but
when the program reaches line 59,000 it crashes and closes.
Sny clue about why this is happening?
Thanks!!
Alfredo Hidalgo-Miranda
Instituto Nacional de Medicina Genómica
Sun Mar 5, 2006 11:14 pm
--- In dc...@yahoogroups.com, ALFREDO HIDALGO <alfhm@...> wrote:
Hi!!
Thank you all for your help, the problem was, as you
suggested, that I had some chips that were not paired,
I have solved this and it runs fine.
Thanks again!!
--- mei xu <mmmxxxuuu2003@...> wrote:
You can put all the files into a database, then
query whatever you want. If you are not familiar
with database and need help, please send your
questions.
Reply all
Reply to author
Forward
0 new messages