[nilshomer]

Hello,

Does the git command above still work? I've tried it a few times today with no luck:

$ git clone git://dnaa.git.sourceforge.net/gitroot/dnaa/dnaa
Initialized empty Git repository in /[my local path]/dnaa/.git/
dnaa.git.sourceforge.net[0: 216.34.181.91]: errno=Connection timed out
fatal: unable to connect a socket (Connection timed out)

Thanks,
Leonardo

PS I'll try later from home as I guess that it could be a local network issue.

Just tested, it works.

[kbushley]

Simulating longer PE Illumina reads,

Hi. I finally found this after searching a bit...on the metasim website there are two configuration files for both 60 and 80 pb PE illumina reads...basically, these contain all the parameters for Illumina PE error models and you can upload it as a configuration file. Hope that helps.


kathryn

[KevinLam]

Ah, get the source code via git as there is not release yet:

git clone git://dnaa.git.sourceforge.net/gitroot/dnaa/dnaa

Nils

is it missing some files?
Just did a git clone
but I can't configure / make


$ ./configure
bash: ./configure: No such file or directory

[nilshomer]

is it missing some files?
Just did a git clone
but I can't configure / make


$ ./configure
bash: ./configure: No such file or directory

try this before configure:

sh autogen.sh

I have updated the INSTALL to include this step. Thanks for spotting the poor documentation.

[KevinLam]

Thanks Nils!
Actually there's only one shell script so it's quite evident (my bad)
anyway i ran that

sh autogen.sh 
Preparing the dnaa build system...please wait

ERROR:  Unable to locate GNU Autoconf.

ERROR:  To prepare the dnaa build system from scratch,
        at least version 2.52 of GNU Autoconf must be installed.


autogen.sh does not need to be run on the same machine that will
run configure or make.  Either the GNU Autotools will need to be installed
or upgraded on this system, or autogen.sh must be run on the source
code on another system and then transferred to here. -- Cheers!

is it possible for you to include the autoconf files?
I do not have that installed on my system

[nilshomer]

Thanks Nils!
Actually there's only one shell script so it's quite evident (my bad)
anyway i ran that

sh autogen.sh 
Preparing the dnaa build system...please wait

ERROR:  Unable to locate GNU Autoconf.

ERROR:  To prepare the dnaa build system from scratch,
        at least version 2.52 of GNU Autoconf must be installed.


autogen.sh does not need to be run on the same machine that will
run configure or make.  Either the GNU Autotools will need to be installed
or upgraded on this system, or autogen.sh must be run on the source
code on another system and then transferred to here. -- Cheers!

is it possible for you to include the autoconf files?
I do not have that installed on my system

Probably not the best idea to include autoconf with source code. If you mean the ./configure script, then it will be included in the releases (no release yet). You will have to either install the appropriate autoconf version or you can PM/email me and I would be happy to send you a tar-ball.

[plichel]

To not to waste threads, does anybody know whether there is a read simulater that can sample also some known snps from, say, a dbsnp file or similar ?
Thanks !

[srividya]

wgsim

Hello,

I am using wgsim to generate simulated reads of 76bp length(Solexa).

The fastq that is generated - Is it solexa fastq or sanger fastq ? Since there is no options to specify the fastq type required, I thought it to be Sanger. Is it correct?

Thanks,
Srividya

[lcollado]

Hello srividya,

I don't know the answer, but you can find out using the ASCII table: http://es.wikipedia.org/wiki/ASCII

Solexa fastq (>= 1.3) won't have any values below 64. Meaning that numbers (48 to 57 in decimal ASCII) shouldn't appear in the quality lines of your fastq file.

Greetings,
Leonardo

[srividya]

Hello,

Thanks for the reply , all the reads had a quality score of 2. Now, I can safely consider them to be Sanger.

Thanks,
Srividya

[lcollado]

No problem and I'm glad you were able to solve your question

Leo

[tldgID]

Questions regarding synthetic data generation

Hi all,

I found this thread about generating synthetic reads for Illumina platform and since I need to generate such synthetic data, I post my question here (as opposed to creating a new thread!).

1) is it possible to generate SE reads and not PE?

2) does anyone know the advantage/disanvantages of “wgsim” from SAMTOOLs vs. “dwgsim” from the DNAA package? What has been modified in dwgsim? it is not very clear to me, since the README file of DNAA package says that:
“This is a fork of the SAMtools wgsim, since certain assumptions are made that we do not agree with.”
what are these assumptions? What has been modified? Is there any publication that elaborates these issues?

3) is there any statistical consideration involved in the generation of the reads? e.g. larger genes on the genome get more reads? Or is there any distribution-related consideration while sheering the reference genome? is the errors distributed uniformly in both software?

4) any other recommendations for synthetic data generation?

Thank you for any help in advance

[nilshomer]

Hi all,

I found this thread about generating synthetic reads for Illumina platform and since I need to generate such synthetic data, I post my question here (as opposed to creating a new thread!).

1) is it possible to generate SE reads and not PE?

2) does anyone know the advantage/disanvantages of “wgsim” from SAMTOOLs vs. “dwgsim” from the DNAA package? What has been modified in dwgsim? it is not very clear to me, since the README file of DNAA package says that:
“This is a fork of the SAMtools wgsim, since certain assumptions are made that we do not agree with.”
what are these assumptions? What has been modified? Is there any publication that elaborates these issues?

3) is there any statistical consideration involved in the generation of the reads? e.g. larger genes on the genome get more reads? Or is there any distribution-related consideration while sheering the reference genome? is the errors distributed uniformly in both software?

4) any other recommendations for synthetic data generation?

Thank you for any help in advance

1) Yes, specify "-2 0".
2) The fork was done to provide better color space (SOLiD) support, in particular to include the first color and adapter.
3) Random read placement, errors distributed according to the error rate.

[tldgID]

1) Yes, specify "-2 0".
2) The fork was done to provide better color space (SOLiD) support, in particular to include the first color and adapter.
3) Random read placement, errors distributed according to the error rate.

Thank you Nils!

About Q2: so, if I need Illumina-like synthetic data, it won't make a difference to use “wgsim” or “dwgsim”?

About Q3: can you elaborate more about “Random read placement”? My understanding is that the error rate is pre-specified, then when the reads are generated, in each position, the nt can be changed according to the error rate. Is this related to “Random read placement” or you meant something else?

Thanks again

[nilshomer]

Q2: there are a number of differences, including left-justification of indels and small bug fixes. You will notice differences and I encourage you test both out as I cannot predict all the differences.

Q3: a read's start position is randomly drawn from all possible start positions. Random errors are then introduced according to the per-base error rate.