Inferred copy number and filtering SNP list file

[Sunny]

Hi, Cheng,

I have some new problems during CN/LOH analysis of my 500K data:
1. After I finished Analysis/Chromosome for unpaired CN/LOH analysis,
I got the raw copy number, but the inferred copy number for all the
tumor samples are "0", what may cause this problem?
2. If I put the filtering SNP list when I do Analysis/Chromosome,
program run much more slowly than using all SNPs. I just put all
selected SNPs in one column in the txt file without column name, is
that a correct format for filtering SNP list file?
3. When I use Analysis/Chromosome after open combined data(Nsp+Sty)
by Analysis/Get external data, I could not select Analysis Method
because this function was disactivated.

Any ideas or suggestions will be greatly appreciated.

Regards,

Sunny

[Cheng Li]

Sunny,

1. At "Analysis/Chromosome/Options", make sure "Inferred copy method"
is not "Do not infer". Or you can paste the output message of the
"Analysis/Chromosome" step here.

2. Is it slow to read the SNP list, or at the display stage? The file
format is correct. If the SNP list is by the same order as those in
the CDF file (or in "Tools/export expression data"), the reading speed
shall be fast.

3. Check "SNP data" at "Analysis/Get external data".

Cheng

[Sunny]

Hi, Cheng,

Thanks for your help. Please see my response below:

> 1. At "Analysis/Chromosome/Options", make sure "Inferred copy method"
> is not "Do not infer". Or you can paste the output message of the
> "Analysis/Chromosome" step here.

I selected "Hidden Markov Model" as Inferred copy method.

The output message is:

{View data along chromosome
Reading genome information file 'K:\Lab Members\Saixia\SNP_Projects
\500K_Cancer\AnalysisBydChip\InfoFiles\hg17\Mapping500K genome info
hg17.txt'...
Probe set 500000
Read human genome information for 500568 probe sets

Reading reference gene file 'K:\Lab Members\Saixia\SNP_Projects
\500K_Cancer\AnalysisBydChip\InfoFiles\hg17\hu refGene hg17
symbol.xls'...
Read refGene entries for 24692 genes

Keeping only one chromosome...

Obtaining SNP or LOH calls...
Unit 40220

Look up sample information file...
Use sample information column 'Ploidy(numeric)'
Use sample information column 'Gender'
No sample information column named 'RefBatch'

Obtaining signal means of each probe set...
Array list file: 'K:\Lab Members\Saixia\SNP_Projects\500K_Cancer
\AnalysisBydChip\Masto_Nsp+Sty_array_list.txt'
Unit 40220
All samples are in one batch
Default batch (170 samples with no 'RefBatch' value specified):
151 normal samples are used to compute signal mean and standard
deviation

Use genotypes of 151 normal samples to obtain genotype dependence
and heterozygosity probablities...
Compute SNP heterozygous rate and dependence probability from
reference normal samples for tumor-only LOH...

Inferring LOH using the 'Hidden Markov Model' method...

Estimated sample LOH rate: 0.02 0.00 0.00 0.00 0.00 0.02 0.00 0.11
0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.04 0.00 0.00 0.00
Chro 1, marker 40001 - 40220, Sample JAR-1-MastoSNP-T
Inferred sample LOH rate: 0.09 0.06 0.06 0.07 0.06 0.10 0.09 0.12
0.07 0.07 0.06 0.07 0.09 0.05 0.13 0.11 0.06 0.06 0.05

Check whether some LOH regions are likely due to haplotype
structure (existing in >10% reference samples)...
Chro 1, JAR-1-MastoSNP-A, at 18Mb, 76 markers, identical with
27% of 151 reference samples, removed
Chro 1, JAR-1-MastoSNP-A, at 19Mb, 67 markers, identical with
49% of 151 reference samples, removed
Chro 1, JAR-1-MastoSNP-A, at 21Mb, 61 markers, identical with
18% of 151 reference samples, removed
Chro 1, JAR-1-MastoSNP-A, at 21Mb, 56 markers, identical with
68% of 151 reference samples, removed
........
Inferring copy number by HMM...
Sample JAR-1-MastoSNP-A
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-B
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-C
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-D
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-E
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-F
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-G
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-H
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-I
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-J
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-L
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-M
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-N
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-O
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-P
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-Q
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-R
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-S
Chro 1, marker 40001 - 40220
Sample JAR-1-MastoSNP-T
Chro 1, marker 40001 - 40220

Finished in 00 hours 00 minutes 42 seconds}

> 2. Is it slow to read the SNP list, or at the display stage? The file
> format is correct. If the SNP list is by the same order as those in
> the CDF file (or in "Tools/export expression data"), the reading speed
> shall be fast.

It is slow to read the SNP list. I'll try resort the SNP list.

> 3. Check "SNP data" at "Analysis/Get external data".

I corrected it now by checking "SNP data". Thanks.

Sunny

>
>
>
> -----Original Message-----
> From: dchip-s...@googlegroups.com
>
> [mailto:dchip-s...@googlegroups.com] On Behalf Of Sunny
> Sent: Tuesday, September 01, 2009 5:05 PM
> To: dChip Software
> Subject: [dchip message 343] Inferred copy number and filtering SNP
> list file
>
> Hi, Cheng,
>
> I have some new problems during CN/LOH analysis of my 500K data:
> 1.  After I finished Analysis/Chromosome for unpaired CN/LOH analysis,
> I got the raw copy number, but the inferred copy number for all the
> tumor samples are "0", what may cause this problem?
> 2. If I put the filtering SNP list when I do Analysis/Chromosome,
> program run much more slowly than using all SNPs.  I just put all
> selected SNPs in one column in the txt file without column name, is
> that a correct format for filtering SNP list file?
> 3.  When I use Analysis/Chromosome after open combined data(Nsp+Sty)
> by Analysis/Get external data, I could not select Analysis Method
> because this function was disactivated.
>
> Any ideas or suggestions will be greatly appreciated.
>
> Regards,
>

> Sunny- Hide quoted text -
>
> - Show quoted text -

[Cheng Li]

Sunny,

Since only chromosome 1 is analyzed (specified at
"Analysis/Chromosome", can you check if chromosome 1 copy number is 0?

Cheng

[Sunny]

Cheng,

I just picked chromosome 1 for testing. The raw copy number for all
SNPs on chr 1 looks right, but the inferred copy number is 0 for all
SNPs on Chr 1. The worse thing is that I also tried chromosome 2
alone and chr 3 alone after I run for chr 1, there were no CN/LOH
results came out.

Sunny

> > - Show quoted text -- Hide quoted text -

[Cheng Li]

Can u send me the screenshot of 'options/chromosome'?

Cheng

在 Sep 1, 2009，9:29 PM，Sunny <yan...@yahoo.com> 写到：

[Cheng Li]

Sunny,

 

Thanks for sharing the solution.

Cheng

---

From: Yans [mailto:yan...@yahoo.com]
Sent: Wednesday, September 02, 2009 1:33 PM
To: Cheng Li
Subject: Re: [dchip message 347] Re: Inferred copy number and filtering SNP list file

 

 

Hi, Cheng,

 

I finally figured out what caused the problem.  My sample information file was created from JMP which uses dot representing missing value, and I forgot to remove the dots in the txt file in the Ploidy column for the tumor samples.  Sorry for wasting your time and thanks for all your help.

 

Regards,

 

Sunny

 

 

Thanks. 

Please see attached file for the screenshots.  BTW, the output message shown that No sample information column named 'RefBatch'.  Is the 'RefBatch' required for the sample information file? Also, for the option/chromosome, I tried checking or unchecking scale copy number mode to 2 copy, the results are same.

 

Sunny

 

 

--- On Tue, 9/1/09, Cheng Li <lch...@gmail.com> wrote:

 

From: Cheng Li <lch...@gmail.com>

Subject: Re: [dchip message 347] Re: Inferred copy number and filtering SNP list file

To: "yan...@yahoo.com" <yan...@yahoo.com>

Cc: "dChip Software" <dchip-s...@googlegroups.com>

Date: Tuesday, September 1, 2009, 10:46 PM