a error

20 views
Skip to first unread message

chunjiang he

unread,
Nov 3, 2009, 2:39:42 PM11/3/09
to alt_pre...@googlegroups.com
Hi,

When I run our exon array data (CEL), I got the error like this. Can anyone help me?


Traceback (most recent call last):
  File "AltAnalyze.py", line 3109, in ?
  File "AltAnalyze.py", line 2726, in AltAnalyzeSetup
  File "AltAnalyze.py", line 2586, in __init__
  File "AltAnalyze.py", line 2983, in AltAnalyzeMain
  File "AltAnalyze.py", line 114, in parse_affymetrix_annotations
  File "BuildAffymetrixAssociations.pyc", line 717, in importAffymetrixAnnotations
  File "BuildAffymetrixAssociations.pyc", line 198, in parse_affymetrix_annotations
UnboundLocalError: local variable 'sp' referenced before assignment

Best,
Jiang

Nathan Salomonis

unread,
Nov 3, 2009, 4:04:29 PM11/3/09
to alt_pre...@googlegroups.com
Hi Jiang,

That is definately a bug (at least it shouldn't terminate like this).
Which specific array are you using? What operating system are you on
(if Mac, Tiger or Leopard)? Did you manually download any affymetrix
annotations files or did you just let AltAnalyze get these for you.
The bug occurs when handling an Affymetrix CSV annotation file with a
format that is different than it expects when adding gene annotations
to the calculated gene expression values prior to alternative splicing
analysis.

Best,
Nathan
> --
>
> You received this message because you are subscribed to the Google Groups
> "Alternative Splicing and Functional Prediction" group.
> To post to this group, send email to alt_pre...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/alt_predictions?hl=en.
>

camelbbs

unread,
Nov 3, 2009, 4:34:28 PM11/3/09
to Alternative Splicing and Functional Prediction
Hi,
I use the GeneChip® Human Exon 1.0 ST Array. And my OS is windows
vista x64. I have downloaded the annotation file HuEx-1_0-st-
v2.na29.hg18.probeset.csv manually. I also run the sample CEL files
and also got this error.

Best,
Jiang

camelbbs

unread,
Nov 3, 2009, 4:43:00 PM11/3/09
to Alternative Splicing and Functional Prediction
Another, the library files are:
HuEx-1_0-st-v2.r2.antigenomic.bgp
HuEx-1_0-st-v2.r2.clf
HuEx-1_0-st-v2.r2.pgf

Nathan Salomonis

unread,
Nov 3, 2009, 4:49:23 PM11/3/09
to alt_pre...@googlegroups.com
Hi Jiang,

AltAnalyze actually doesn't require any annotation files for the exon
arrays, so I would just move this out of the program's space and try
again. I realized recently that users may wish to try include this
file and the next version will just ignore it. So if in one of the
AltDatabase/Affymetrix directories, just take it out and run again.

Best,
Nathan

Nathan Salomonis

unread,
Nov 3, 2009, 4:50:19 PM11/3/09
to alt_pre...@googlegroups.com
By the way, AltAnalyze doesn't need you to supply the library files.
It can automatically grab this after determing which array type you
are using.

Best,
Nathan

Nathan Salomonis

unread,
Nov 3, 2009, 5:00:15 PM11/3/09
to alt_pre...@googlegroups.com
As another note I highly recommend loading the AltAnalyze DomainGraph
results in DomainGraph when finished. These will help visualize the
predictions in a biological context.
Best,
Nathan

On Nov 3, 2009, at 1:50 PM, Nathan Salomonis <nsalo...@gmail.com>

camelbbs

unread,
Nov 3, 2009, 5:16:57 PM11/3/09
to Alternative Splicing and Functional Prediction
Hi,
Thanks very much.
When I remove the HuEx-1_0-st-v2.na29.hg18.probeset.csv, It is ok.
So I also find two annotation files HuEx-1_0-st-
v2.na29.hg18.transcript.csv and HuEx-1_0-st-
v2.na28.hg18.transcript.csv.
Which version I can use? Or I can use both.

Best,
Jiang

On Nov 3, 4:00 pm, Nathan Salomonis <nsalomo...@gladstone.ucsf.edu>
wrote:
> As another note I highly recommend loading the AltAnalyze DomainGraph  
> results in DomainGraph when finished. These will help visualize the  
> predictions in a biological context.
> Best,
> Nathan
>
> On Nov 3, 2009, at 1:50 PM, Nathan Salomonis <nsalomo...@gmail.com>  
> wrote:
>
>
>
> > By the way, AltAnalyze doesn't need you to supply the library files.
> > It can automatically grab this after determing which array type you
> > are using.
>
> > Best,
> > Nathan
>
> > On Tue, Nov 3, 2009 at 2:49 PM, Nathan Salomonis  
> > <nsalomo...@gmail.com> wrote:
> >> Hi Jiang,
>
> >> AltAnalyze actually doesn't require any annotation files for the exon
> >> arrays, so I would just move this out of the program's space and try
> >> again. I realized recently that users may wish to try include this
> >> file and the next version will just ignore it. So if in one of the
> >> AltDatabase/Affymetrix directories, just take it out and run again.
>
> >> Best,
> >> Nathan
>

camelbbs

unread,
Nov 3, 2009, 5:22:06 PM11/3/09
to Alternative Splicing and Functional Prediction
Hi,
I am sorry I write the wrong filenames. I mean that "HuGene-1_0-st-
v1.na29.hg18.transcript.csv" and "HuGene-1_0-st-
v1.na28.hg18.transcript.csv" are same?

I also have a question. If I don't use the exon array annotation, does
the AltAnalyze can automated get the exon probe-set informations?

Best,
Jiang

Nathan Salomonis

unread,
Nov 3, 2009, 5:31:04 PM11/3/09
to alt_pre...@googlegroups.com
For exon arrays, most of the necessary annotations are in the existing
database, it actually doesn't need and of the exon array annotation
files. If you want GO annotations added though, you can get these from
non-exon arrays. For this You have two options:
1) let AltAnalyze download it for you when prompted
2) download the Gene 1.0 transcript annotations and choose these.

See the tutorial (tutorial 2) linked to from the AltAnalyze download
page for more information.

Best,
Nathan

Sent from my iPhone

Nathan Salomonis

unread,
Nov 3, 2009, 5:34:06 PM11/3/09
to alt_pre...@googlegroups.com, Alternative Splicing and Functional Prediction
AltAnalyze only needs CEL files. It has or will download anything else
It needs.
Best,
Nathan

camelbbs

unread,
Nov 3, 2009, 6:38:34 PM11/3/09
to Alternative Splicing and Functional Prediction
Dear Nathan,
Thanks very much. It is a very beautiful software. I am using it and
if i have any questions I will ask you.
Best,
Jiang

On Nov 3, 4:34 pm, Nathan Salomonis <nsalomo...@gladstone.ucsf.edu>
wrote:
> AltAnalyze only needs CEL files. It has or will download anything else  
> It needs.
> Best,
> Nathan
>

camelbbs

unread,
Nov 4, 2009, 12:21:44 PM11/4/09
to Alternative Splicing and Functional Prediction
Hi Nathan,
I found the library files the software used is based on v2: HuEx-1_0-
st-v2.r2.clf. But the csv files is based on v1: HuGene-1_0-st-
v1.na28.hg18.transcript.csv. So is it OK?
Another, I found some targeted miRNAs are not included in the
database. For example, the targeted miRNAs in Hoxa9. So if I want to
add more targeted miRNAs into the database, can I change the database
and how? I found the file Hs_probeset_microRNAs_any.txt, but it is
based on the sequence. So the software detect the miRNA according to
the sequence?

Thanks so much.
Best,
Jiang

On Nov 3, 4:34 pm, Nathan Salomonis <nsalomo...@gladstone.ucsf.edu>
wrote:
> AltAnalyze only needs CEL files. It has or will download anything else  
> It needs.
> Best,
> Nathan
>

camelbbs

unread,
Nov 4, 2009, 12:37:55 PM11/4/09
to Alternative Splicing and Functional Prediction
Another, if I have 3 groups and I want to compare group1 to other 2
groups can I do it? I see I can choose the disease groups and the
control groups. But can the control groups include more than 1 groups?
Thanks,
Best,
Jiang

On Nov 3, 4:34 pm, Nathan Salomonis <nsalomo...@gladstone.ucsf.edu>
wrote:
> AltAnalyze only needs CEL files. It has or will download anything else  
> It needs.
> Best,
> Nathan
>

Nathan Salomonis

unread,
Nov 4, 2009, 1:02:57 PM11/4/09
to alt_pre...@googlegroups.com, camelbbs
Hi Jiang,

See below:

"I found the library files the software used is based on
v2: HuEx-1_0-
st-v2.r2.clf. But the csv files is based on v1:
HuGene-1_0-st-
v1.na28.hg18.transcript.csv. So is it OK?"

>>> Yes, this is normal... The only information that AltAnalyze uses for exon arrays from the HuGene-1_0-st-v1.na28.hg18.transcript.csv is Ensembl gene to Gene Ontology annotations. AltAnalyze can get also get this from any non-exon array.

"Another, I found some targeted miRNAs are not included in
the
database. For example, the targeted miRNAs in Hoxa9. So if
I want to
add more targeted miRNAs into the database, can I change
the database
and how? I found the file Hs_probeset_microRNAs_any.txt,
but it is
based on the sequence. So the software detect the miRNA
according to
the sequence?"

>>> Good question. You found the right file. AltAnalyze_v1beta/AltDatabase/Hs/exon/Hs_probeset_microRNAs_any.txt contains all microRNA target sequences from MirBase, TargetScan, Miranda and PicTar that could be first matched to the correct gene. If Ensembl version 51 IDs could not be mapped to the gene IDs given in these databases or the annotation you are looking for is missing, then it won't be found in this table. However, if you have mapping you think are valid, between probeset and microRNA, you can add or replace entries in the above table. This table has 4 fields: probeset, miRNA, miRNAseq, and database origin. If you add your own entries, each line must have four entries, but you can leave the last two field as what ever you want. The program will likely give an exception if there are NOT four fields here. For example, If you want to add the relationships probeset 1234 to miRNA-1, you would add the line:
1234 miRNA-1 null null.

Best,
Nathan

Nathan Salomonis

unread,
Nov 4, 2009, 1:07:45 PM11/4/09
to alt_pre...@googlegroups.com, camelbbs
Hi Jiang,

We are working on adding new alternative exon statistical
methods to AltAnalyze, but in the meantime AltAnalyze is
restricted to just two comparison groups. There are ways
to compare multiple groups that involve AltAnalyze and
Bioconductor though. This is something we did in a recent
paper that is in press. Let's discuss this more offline.

Best,
Nathan
Reply all
Reply to author
Forward
0 new messages