[nilshomer]

Hi Nils,
just stumbled across dwgsim in my search for a read simulator, love the fact that I can generate Illumina, Solid and Ion-torrent on the same tool as I have data from all platforms. Nice touch as well on including the galaxy wrappers
At the risk of sounding ungrateful, is there an ETA on the dwgsim_pileup_eval.pl? You'd solve all my problems in one go.
Thanks for sharing,
Cheers
Seb

No, I haven't thought about when/if I am going to update it. It works fine on samtools pileup data, but now needs VCF support. I'd also like to re-write it in python to improve maintainability. Nevertheless, I am always be happy to accept patches or contributions, as your free time is as free as mine .

[NextGenSeb]

Fair enough, maybe that gives me teh motivation to finally improve my almost non-existent python skills

Got a few more questions fo ryou though:

1. The "bed-like" format for candidate mutations (-b option), what is that supposed to look like? From the header files the first three columns are pretty clear, but I am stuck with regards to the mutation type and the rest of the file

2. Still in reagards to mutations, is there a way to simulate specific mutation frequencies? I would like like to test aligners and variant callers for somatic mutations and hence need a degree of frequency control over the inserted mutations.

3. The IonTorrent homopolymer errors (elegedly) got better with the new chemistry. Is there a way to potentially adjust the error frequencies in the program for this?

Thanks for the quick reply and your comments.

Cheers
Seb

[nilshomer]

Fair enough, maybe that gives me teh motivation to finally improve my almost non-existent python skills

Got a few more questions fo ryou though:

I found those skills come in handy!

1. The "bed-like" format for candidate mutations (-b option), what is that supposed to look like? From the header files the first three columns are pretty clear, but I am stuck with regards to the mutation type and the rest of the file

Use the VCF support in the latest GIT code. If you still want to use the bed-like format, the source code is the documentation!

2. Still in reagards to mutations, is there a way to simulate specific mutation frequencies? I would like like to test aligners and variant callers for somatic mutations and hence need a degree of frequency control over the inserted mutations.

None yet, but I don't see why that couldn't be added (search for instances of 0.5 in the code).

3. The IonTorrent homopolymer errors (elegedly) got better with the new chemistry. Is there a way to potentially adjust the error frequencies in the program for this?
Seb

Yes, see the "-e/-E" options. I like the "elegedly" .

[NextGenSeb]

Great, thanks again for your help. I'll give it a go and let you know the outcome. Once I have results from the new chemistry I'' try to remember to put that up as well.

Cheers
Seb

[chintanspy]

Hi Nils,

what parameters are used for placing a read under 'mi' category?

After observing the results, not all the reads which map at different co-ordinate than simulated position are categorized under 'mi'.

[nilshomer]

Hi Nils,

what parameters are used for placing a read under 'mi' category?

After observing the results, not all the reads which map at different co-ordinate than simulated position are categorized under 'mi'.

From the header: "# mi' | the number of reads mapped incorrectly that should be mapped be mapped at or greater than that threshold".

It can also be affected by the "-g" parameter.

[chintanspy]

It can also be affected by the "-g" parameter.

I could not find " -g" parameter. Am I looking at wrong patch (nh13-DWGSIM-2fc8222) ?

[nilshomer]

I could not find " -g" parameter. Am I looking at wrong patch (nh13-DWGSIM-2fc8222) ?

See "dwgsim_eval".

[plantae]

I was confused by "-d" opinition
NB: the -d option was previously incorrectly stated as being the outer distance, but is in fact the inner distance.

for a PE-read

5--->3______3<---5

"-d" means for distance between 3' and 3' of the PE reads?

[nilshomer]

I was confused by "-d" opinition
NB: the -d option was previously incorrectly stated as being the outer distance, but is in fact the inner distance.

for a PE-read

5--->3______3<---5

"-d" means for distance between 3' and 3' of the PE reads?

See the help message for the latest version.

[adaptivegenome]

Are you planning to publish this? We just cited it in a paper we submitted, but since there was no paper we cited the link to the software....

[plantae]

See the help message for the latest version.

I have intalled version 0.1.10, the help message did not clear specify the "-d" opinion

-d INT inner distance between the two ends [500]

"the two ends"
means for two pair-end reads?
or 3' end of two pair-end reads?
or 5' end of two pair-end reads?

[liu_xt005]

I have used dwgsim 0.1.10 to generate reads from hg19 chromosome 22, and tested the sensitivities of variant callers SAMtools, GATK, and glfSingle.
The default per base error rate 0.02 is used.
Even up to 100X ('-C 100' in simulation), GATK and glfSingle can identify only 86% the the variants. But SAMtools can call 95%.
Shall I reduce the error rate?

[nilshomer]

Please post your questions in a new thread, thanks!