dChip for Affy promoter arrays

[Cheng Li]

--- Forwarded old messages ---

Posted: 23 May 2007 11:10 pm Post subject: dChip for Affy promoter
arrays

--------------------------------------------------------------------------------
Wed Mar 21, 2007 2:10 am
--- In dc...@yahoogroups.com, "Lavinia" <lavinia.gordon@...> wrote:

Hi

I am trying to use dChip with Human Promoter Arrays 1.0

I have obtained the BPMAP files from the Affymetrix website for the
Human tiling 1.0R Promoter arrays.
I have downloaded the ProbeExporter software
I ran ProbeExporter to convert the two bpmap files:
Hs_PromPF_v02-3_NCBIv34.bpmap
Hs_PromPR_v02-3_NCBIv34.bpmap
to text format and renamed *.bpmap.txt to *.bpmap

At dChip "Open group", I specified the data directory as the one
containing the tiling array CEL files, and specified "Other
information/CDF file" as Hs_PromPF_v02-3_NCBIv34.bpmap (converted
text
format). I specified that it was a 'tiling array'. A genome info file
was created and the bmap file was read but it fails with:
Error: CheckCellPos() failed at probe set chr1_17101, probe pair 1.
Is the SNP array type specified in dChip, or does this probe cell
occur in two probe sets?
Opening group not completed}

with regards

Lavinia Gordon.


pipibao10 Posted: 28 Aug 2007 06:22 pm Post subject:

--------------------------------------------------------------------------------
Hi Lavinia:

I was working on Affymetrix promoter array and got the same error
message in the beginning. However, I specified the file for "Other
information/CDF file" as Hs_PromPR_v02-3_NCBIv34.bpmap, instead of
Hs_PromPF_v02-3_NCBIv34.bpmap (as you did), then it worked (basically
changed the annotation file from "PF -- forward strand" to "PR --
reverse stran").

Hope this will work for you, too! Good luck!


lch Posted: 29 Aug 2007 04:23 pm Post subject:

--------------------------------------------------------------------------------
Hi pipibao10,

Thanks for sharing your experiences. I also moved other related
messages from Yahoo group below.

Mon Apr 30, 2007 7:50 pm
Re: dChip for Affy promoter arrays.
Luke,

You may open a CEL file by Wordpad (ignore warning) and check if the
header sections contains Hs_PromPF_v02-3_NCBIv34.1sq (to match bpmap
file name). If there is difference, rename bpmap file to match header
section.

Cheng

--- In dc...@yahoogroups.com, Luke <lukelan.tw@...> wrote:

Hi Dr.Li,
Thanks for your reply. I checked "Open group/options/Array has only PM
probe" and got another msg "Found no files of this type; are correct
data directory..". I put the screenshot in the attached file.
Thanks.
Luke


Sun Apr 22, 2007 5:25 pm
Re: dChip for Affy promoter arrays.

Luke,

Prbably "Open group/options/Array has only PM probe" should be
checked.

Cheng

--- In dc...@yahoogroups.com, "Look" <lukelan.tw@...> wrote:

Hi Dr. Li,
I have the same question too.
Would you please help to instruct how to analyze the promoter data?
Thanks.
Luke


lch Posted: 05 Sep 2007 06:00 pm Post subject:

--------------------------------------------------------------------------------
Bao-Li,

1. You may try to open the file in Excel and delete the last empty
column (may contain tabs) and re-save. If not working you can send me
the file.

2. You may try “Analysis/compare samples” to get a list of probe sets
and specify it at “Analysis/Chromosome”. But with 1 array of each you
should specify large fold change and mean difference to reduce noise.

Cheng


From: Baoli Chang
Sent: Thursday, August 30, 2007 6:44 PM
To: Cheng Li
Subject: questions for analyzing Affymetrix promoter array using
dCHIP

Hi Dr. Li:

We have this email conversion last week, and finally I found what I
did that made dCHIP not working (I have posted my experiences on the
dCHIP forum).

I have two more questions for you:

1) I ran the expression data (output from dCHIP) through ComBat, and
try to see whether I can use dCHIP to do further analysis. However, I
can’t make dCHIP to read in the external file correctly (in the “get
external data” setting, I kept “skip columns 2 to 1 and skip rows 2 to
1”). However, it always read one less array (which I guess equals to
reading one less column). I can’t figure out why.
2) Is there any high level analysis in dCHIP you think is reasonable
for analyzing titling array data? I tried to compare the CHIP-on-chip
data between paired tumor and normal sample (1 array each for tumor
and normal).

Thanks very much for your help!

Bao-Li Chang, PhD