Simulated reads mapping to same region - maq simulate & dwgsim

gprakhar · 02-18-2011, 11:40 AM

Hello,

I have written a CNV Simulator, which takes fasta(eg. chr22 from hg18, repeat masked, all 'N' removed) file as input and inserts CNV at random locations.
I then generate reads using maq simulate.
To achieve a coverage of around 20x.
I give,
maq simulate -d 200 -N 2466490 -1 75 -2 75 sim_read1.out sim_read2.out chr22_sim_ref_fi.fa simupar.dat

After generation of reads and Alignment by BFAST, to verify I wrote a perl script to bin the sam file into 5kb windows and calculate how many reads map to each window.
I find that most(>95%) of the read are mapping to only the last window.

I read the man page, still not very clear about few issues.
If I specify the parameters for simulate, is it necessary to provide the .dat file?
If yes, then what kind of fastq file should I train simutrain on?
Also the man page mentions that there is a .dat file available for download, but I am not able to find it.
I have also tried with dwgsim now, the same problem occurs, all reads mapping to the last window.

Can anyone point out what I am doing wrong.

Thanking you,
Prakhar Gaur

gprakhar · 02-18-2011, 03:26 PM

Tried with chr18.fa from hg18.
I am getting the same results, almost all reads mapping to the same region.

$dwgsim -d 200 -1 75 -2 75 chr18.fa out.bwa.read1.fq out.bwa.read2.fq out.bfast.fq
then used Bfast.

gprakhar · 02-18-2011, 11:12 PM

There is No issue with maq simulate or dwgsim.
It was my mistake which caused the problem.
The Perl pipeline was faulty.

Sorry for any inconvenience caused.
--
Prakhar Gaur

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Using MAQ to simulate illumina pair-end reads	yjx1217	Bioinformatics	0	10-27-2011 08:38 AM
dwgsim to simulate Illumina reads	gene coder	Bioinformatics	1	07-07-2011 09:51 AM
MAQ simulated read header names	MBekritsky	Bioinformatics	0	11-18-2010 06:46 PM
Too many reads mapping towards intronic region	sanush	SOLiD	6	04-14-2010 03:23 AM
Too many reads mapping towards intronic region	sanush	RNA Sequencing	1	04-13-2010 10:10 AM

02-18-2011, 11:40 AM	#1
gprakhar Member Location: Bangalore, India Join Date: Aug 2010 Posts: 67	Simulated reads mapping to same region - maq simulate & dwgsim Hello, I have written a CNV Simulator, which takes fasta(eg. chr22 from hg18, repeat masked, all 'N' removed) file as input and inserts CNV at random locations. I then generate reads using maq simulate. To achieve a coverage of around 20x. I give, maq simulate -d 200 -N 2466490 -1 75 -2 75 sim_read1.out sim_read2.out chr22_sim_ref_fi.fa simupar.dat After generation of reads and Alignment by BFAST, to verify I wrote a perl script to bin the sam file into 5kb windows and calculate how many reads map to each window. I find that most(>95%) of the read are mapping to only the last window. I read the man page, still not very clear about few issues. If I specify the parameters for simulate, is it necessary to provide the .dat file? If yes, then what kind of fastq file should I train simutrain on? Also the man page mentions that there is a .dat file available for download, but I am not able to find it. I have also tried with dwgsim now, the same problem occurs, all reads mapping to the last window. Can anyone point out what I am doing wrong. Thanking you, Prakhar Gaur

02-18-2011, 03:26 PM	#2
gprakhar Member Location: Bangalore, India Join Date: Aug 2010 Posts: 67	Tried with chr18.fa from hg18. I am getting the same results, almost all reads mapping to the same region. $dwgsim -d 200 -1 75 -2 75 chr18.fa out.bwa.read1.fq out.bwa.read2.fq out.bfast.fq then used Bfast.

02-18-2011, 11:12 PM	#3
gprakhar Member Location: Bangalore, India Join Date: Aug 2010 Posts: 67	There is No issue with maq simulate or dwgsim. It was my mistake which caused the problem. The Perl pipeline was faulty. Sorry for any inconvenience caused. -- Prakhar Gaur