SEQanswers

Go Back   SEQanswers > Applications Forums > De novo discovery



Similar Threads
Thread Thread Starter Forum Replies Last Post
Acquiring contigs, de Bruijn graphs, velvet bioinf Bioinformatics 8 10-14-2012 07:27 PM
problem for velvet parameters and results hequn Bioinformatics 2 11-17-2011 04:56 AM
velvet assembler contigs into gbrowse Zimbobo Bioinformatics 0 04-15-2010 01:36 PM
number of contigs in velvet bioenvisage Bioinformatics 6 03-24-2010 08:10 PM
BWA and Velvet : Anamolies in contigs and read mapping to reference apratap Bioinformatics 1 01-29-2010 12:44 PM

Reply
 
Thread Tools
Old 06-22-2009, 04:22 AM   #1
kim
Junior Member
 
Location: Korea

Join Date: Jun 2009
Posts: 2
Default Using Velvet options and the results, examine contigs

Hi,

I'm working sequence analysis, and just a few days ago did assembly of short read using Velet.
Okay, I can say I just began.
Whatever, Reads are from Solexa, paired-end, 76 bases in length.

I made various cases by options such as seed length, min contig length, cov cutoff, exp cov, etc.
Among them, seed length and cov cutoff seemed to have critical effect on assembly result.
Min contig length had also some,but exp cov looked not that notable, in my humble sight.

Seed length is quite expected to do so, but I have no idea what was going on with 'cov cutoff'.
With other options fixed, no value given to it, 2 given, 5 given, and 10 made result of considerable difference.
As larger it was, the number of contigs decreased and the size of those dramatically increased.

Therefore, I wanted to know how many reads contributed into contigs, that is, how many reads made those.
But I can't find what file or option to refer.

Additionally, I also wonder how the option 'cov cutoff' made those huge contigs.
Actually, no value (default) and 5 showed average size of contigs differ more than ten times.
Contigs were constructed less than ten percent. (Contig length sum was not that different.)
Finally, are the larger contigs reliable indeed?

Regards,
Kim
kim is offline   Reply With Quote
Old 06-22-2009, 06:55 AM   #2
bioinfosm
Senior Member
 
Location: USA

Join Date: Jan 2008
Posts: 482
Default

you may try the velvetg command with options '-amos_file yes -read_trkg yes' that creates a velvet_asm.afg file that contains some of that information...
bioinfosm is offline   Reply With Quote
Old 06-22-2009, 06:38 PM   #3
kim
Junior Member
 
Location: Korea

Join Date: Jun 2009
Posts: 2
Default Options

thanks.

I already had used '-read_trkg yes' option, but I wasn't able to understand what was changed by that one,
and it seemed there's no output file newly created.

And, I added '-amos_file yes' as your comment. That made a new file named 'velvet_asm.afg'.
Below is the beginning of it :
-------------velvet_asm.afg---------------
{LIB
iid:1
{DST
mea:181615104
std:-9223372036854775808
}
}
{LIB
iid:2
{DST
mea:181615104
std:-9223372036854775808
}
}
{LIB
iid:3
{DST
mea:96
std:1504
}
}
{RED
iid:1
eid:1
seq:
ATGAGCTTCCTTTGATCTCTAGCTTTCCCGATGGCTATGATCCCGTTCCAACACCTCGTG
ATGATAATTCATATAC
.
qlt:
ATGAGCTTCCTTTGATCTCTAGCTTTCCCGATGGCTATGATCCCGTTCCAACACCTCGTG
ATGATAATTCATATAC
.
}
(.............)
--------------------------------

Now I'm trying to use 'amos2ace', generating ACE file from AMOS file (.afg).
Then I can find out the reads in each contig.
If anyone knows another way to identify that information, please inform me.

Anyway, thanks again to bioinfosm.
kim is offline   Reply With Quote
Old 06-25-2009, 07:57 AM   #4
bioinfosm
Senior Member
 
Location: USA

Join Date: Jan 2008
Posts: 482
Default

If you scroll further down in the afg file, there are lines beginning with
src:
These are the read IDs contained in that contig. The ID comes from velvet's Sequences file.

Could you share how amo2ace helped you..
bioinfosm is offline   Reply With Quote
Old 06-30-2009, 06:02 AM   #5
Zigster
(Jeremy Leipzig)
 
Location: Philadelphia, PA

Join Date: May 2009
Posts: 116
Default

when I run kmer/cvCut permutations I gzip the afg files and compile statistics on them later:

Code:
open ASM,"<:gzip","$rootDir/velvet_asm.afg.gz" or die "can't find .afg file";
while(<ASM>){
    if(/\{RED/){$reads++}
    elsif(/\{TLE/){$tiles++}
}
Zigster is offline   Reply With Quote
Old 03-14-2011, 09:10 PM   #6
waterboy
Member
 
Location: India

Join Date: Oct 2010
Posts: 14
Default

Hello,
I am doing transcriptome de novo assembly on SOLiD data, for that I am running velvet and oases pipeline, in the assembly outoff 31 million reads only 1 million reads taking part in assembly!! what is the reason behind that??
Your answers and help would be appreciated.
waterboy is offline   Reply With Quote
Old 03-15-2011, 01:39 AM   #7
Thorondor
Member
 
Location: Heidelberg

Join Date: Feb 2011
Posts: 69
Default

woah, old thread. ;-)

how long are you reads? did you trim before running velvet and oases? which kmers you choose, take care that you do not choose a higher kmer than your minimum read length. Could it be, that the read quality is messed up?
Thorondor is offline   Reply With Quote
Old 03-15-2011, 09:01 PM   #8
waterboy
Member
 
Location: India

Join Date: Oct 2010
Posts: 14
Default

The read length is 50bp. yes I trimmed the reads before running velvet and oases.
The kmer length that we prefer is 25-31. Even after doing that only 2806127 reads out of 31405877 filtered reads participate in the
assembly process. Well coming on to the qaulaity of reads we also performed quality filteration. Hope that covers everything what you asked. But I am still surprised why I am not able to assemble large number of reads from my input data. Could you please provide a robust elaboration on it??

Last edited by waterboy; 03-15-2011 at 09:05 PM.
waterboy is offline   Reply With Quote
Old 03-16-2011, 04:34 AM   #9
Thorondor
Member
 
Location: Heidelberg

Join Date: Feb 2011
Posts: 69
Default

it is really hard to guess.

You trimmed all reads to the same length of 50bp?
What total coverage for your transcriptome do you expect with this 31405877 reads. Only option you have left so far is try even lower kmers and a really low cov_cutoff (e.g. 2).
Thorondor is offline   Reply With Quote
Old 08-18-2011, 12:57 AM   #10
mghita
Member
 
Location: Cambridge

Join Date: Aug 2011
Posts: 10
Default

Does anyone know how to allow for SNPs when assembling sequences with Velvet?

I have downloaded Velvet 1.1.05 and obtained some contigs using my reads, but they don't have SNPs and I am quite sure they should.

Thanks
mghita is offline   Reply With Quote
Old 10-06-2011, 12:26 AM   #11
ramadatta.88
Junior Member
 
Location: singapore

Join Date: Sep 2011
Posts: 7
Default

Getting number of reads in each contig:

I used contrib/extraContigReads.pl in velvet to make contig file and its corresponding reads. After making all the contig files (contig_1_reads.txt,contig_1_reads.txt,........), i just run

grep -c '>' contig_*.txt > Outputfile.txt

The Output file contain contignumber and its corresponding reads..

Last edited by ramadatta.88; 10-06-2011 at 12:28 AM. Reason: assigning title
ramadatta.88 is offline   Reply With Quote
Old 11-10-2011, 08:34 PM   #12
icebreaker
Junior Member
 
Location: india

Join Date: Oct 2011
Posts: 5
Default Getting number of reads in each contig:

Quote:
Getting number of reads in each contig:

I used contrib/extraContigReads.pl in velvet to make contig file and its corresponding reads. After making all the contig files (contig_1_reads.txt,contig_1_reads.txt,........), i just run

grep -c '>' contig_*.txt > Outputfile.txt
That would definately take ages when you are dealing with millions of reads..Best thing would be using AMOS package.

convert the afg file output by velvet to a bnk file

~/src/amos-3.0.0/src/Bank/bank-transact -m my.afg -b my.bnk -c

run depth-of-coverage output

~/src/amos-3.0.0/src/Validation/analyze-read-depth my.bnk -d -r > Outputfile

The output file contains the contigs ID, Corresponding reads, depth of contig. etc..Hope this helps someone
icebreaker is offline   Reply With Quote
Old 02-24-2012, 04:03 AM   #13
rahularjun86
Member
 
Location: Frankfurt(M), Germany

Join Date: Jan 2011
Posts: 58
Default

Dear all,

Following is the velvet_asm.afg file generated by velevt, But I want to have my Read Id's from the Sequence file here in this format. What Id's should I change?? both iid and eid or what??

Thanks,
RAhul

-------------velvet_asm.afg---------------
{LIB
iid:1
{DST
mea:181615104
std:-9223372036854775808
}
}
{LIB
iid:2
{DST
mea:181615104
std:-9223372036854775808
}
}
{LIB
iid:3
{DST
mea:96
std:1504
}
}
{RED
iid:1
eid:1
seq:
ATGAGCTTCCTTTGATCTCTAGCTTTCCCGATGGCTATGATCCCGTTCCAACACCTCGTG
ATGATAATTCATATAC
.
qlt:
ATGAGCTTCCTTTGATCTCTAGCTTTCCCGATGGCTATGATCCCGTTCCAACACCTCGTG
ATGATAATTCATATAC
.
}
(.............)
--------------------------------
__________________
Rahul Sharma,
Ph.D
Frankfurt am Main, Germany
rahularjun86 is offline   Reply With Quote
Old 02-25-2012, 10:52 AM   #14
SES
Senior Member
 
Location: Vancouver, BC

Join Date: Mar 2010
Posts: 275
Default

Quote:
Originally Posted by rahularjun86 View Post
Dear all,

Following is the velvet_asm.afg file generated by velevt, But I want to have my Read Id's from the Sequence file here in this format. What Id's should I change?? both iid and eid or what??

Thanks,
RAhul

-------------velvet_asm.afg---------------
{LIB
iid:1
{DST
mea:181615104
std:-9223372036854775808
}
}
{LIB
iid:2
{DST
mea:181615104
std:-9223372036854775808
}
}
{LIB
iid:3
{DST
mea:96
std:1504
}
}
{RED
iid:1
eid:1
seq:
ATGAGCTTCCTTTGATCTCTAGCTTTCCCGATGGCTATGATCCCGTTCCAACACCTCGTG
ATGATAATTCATATAC
.
qlt:
ATGAGCTTCCTTTGATCTCTAGCTTTCCCGATGGCTATGATCCCGTTCCAACACCTCGTG
ATGATAATTCATATAC
.
}
(.............)
--------------------------------
I would change the "eid" if you must have your own IDs in that file. The "iid" is an internal ID, and my guess is that changing the form of this might cause problems with software designed for reading afg files.
SES is offline   Reply With Quote
Old 06-14-2013, 06:49 AM   #15
flacchy
Member
 
Location: UK

Join Date: Apr 2013
Posts: 33
Default Velvet: how many reads into each contig

Hi, I guess this is the right place to ask...

I am trying to figure out how many reads made it into each contig after I run velvet. I used the flag -read_trkg yes and -amos_file yes.
I found some advise online on how to process them but I am a bit stuck.
So from the amos file I have tried this path:

$ /home/admin/amos-3.1.0/src/Bank/bank-transact -m velvet_asm.afg -b velvet_asm.bnk -c

and then
$/home/admin/amos-3.1.0/src/Validation/analyze-read-depth velvet_asm.bnk -i -d -r > output

The problem is that the file output is something like this :
1 1646 1340 1340 108.275 108.275
2 1 97 97 1.03093 1.03093
3 2085 1589 1589 113.904 113.904
4 175 75 75 219.547 219.547
5 187 162 162 105.944 105.944
6 124 76 76 157.171 157.171
7 1023 768 768 112.914 112.914
8 9 99 99 9.09091 9.09091
9 151 168 168 72.8214 72.8214
10 2224 1602 1602 123.306 123.306
11 6582 5312 5312 107.116 107.116
12 2850 2487 2487 98.653 98.653
13 3 105 105 2.62857 2.62857
14 3 103 103 2.91262 2.91262
15 4350 3350 3350 115.28 115.28
16 1788 1268 1268 128.341 128.341
17 1746 1325 1325 116.024 116.024

could anyone tell me what the filed are or if there is a different way to get the information I need??

Thank you,

F.
flacchy is offline   Reply With Quote
Old 01-16-2014, 05:46 AM   #16
Lszar
Junior Member
 
Location: Brazil

Join Date: Jan 2013
Posts: 2
Default

Quote:
Originally Posted by icebreaker View Post
That would definately take ages when you are dealing with millions of reads..Best thing would be using AMOS package.

convert the afg file output by velvet to a bnk file

~/src/amos-3.0.0/src/Bank/bank-transact -m my.afg -b my.bnk -c

run depth-of-coverage output

~/src/amos-3.0.0/src/Validation/analyze-read-depth my.bnk -d -r > Outputfile

The output file contains the contigs ID, Corresponding reads, depth of contig. etc..Hope this helps someone
Hi Icebreaker,

I'm trying to do exactly what suggested, but I'm unsuccessful. After both commands the result was a directory/, and the output file have this appearance:

1 23 74 74 38.5676 38.5676
2 71 54 54 170.148 170.148
3 49 50 50 136.52 136.52
4 89 54 54 223.463 223.463
5 45 64 64 98.9375 98.9375
6 3 141 141 3.21277 3.21277
7 1 108 108 1.39815 1.39815
8 20 64 64 42.6719 42.6719
9 62 54 54 158.611 158.611
10 10 64 64 22.4375 22.4375
11 37 61 61 77.1639 77.1639
12 32 242 242 17.0579 17.0579
13 20 57 57 51.0175 51.0175

I can't understant this file and I don't see contigs ID, Corresponding reads, depth of contig. etc, as you explained.
Can you help me?
Thanks
Lszar
Lszar is offline   Reply With Quote
Old 01-16-2014, 05:49 AM   #17
Lszar
Junior Member
 
Location: Brazil

Join Date: Jan 2013
Posts: 2
Default

Quote:
Originally Posted by flacchy View Post
Hi, I guess this is the right place to ask...

I am trying to figure out how many reads made it into each contig after I run velvet. I used the flag -read_trkg yes and -amos_file yes.
I found some advise online on how to process them but I am a bit stuck.
So from the amos file I have tried this path:

$ /home/admin/amos-3.1.0/src/Bank/bank-transact -m velvet_asm.afg -b velvet_asm.bnk -c

and then
$/home/admin/amos-3.1.0/src/Validation/analyze-read-depth velvet_asm.bnk -i -d -r > output

The problem is that the file output is something like this :
1 1646 1340 1340 108.275 108.275
2 1 97 97 1.03093 1.03093
3 2085 1589 1589 113.904 113.904
4 175 75 75 219.547 219.547
5 187 162 162 105.944 105.944
6 124 76 76 157.171 157.171
7 1023 768 768 112.914 112.914
8 9 99 99 9.09091 9.09091
9 151 168 168 72.8214 72.8214
10 2224 1602 1602 123.306 123.306
11 6582 5312 5312 107.116 107.116
12 2850 2487 2487 98.653 98.653
13 3 105 105 2.62857 2.62857
14 3 103 103 2.91262 2.91262
15 4350 3350 3350 115.28 115.28
16 1788 1268 1268 128.341 128.341
17 1746 1325 1325 116.024 116.024

could anyone tell me what the filed are or if there is a different way to get the information I need??

Thank you,

F.
Hi Flacchy, do you solve your questions?
I'm having the same problem now, that you have time, time ago.
If you have some solution, please, let me know.
Thanks
Lszar
Lszar is offline   Reply With Quote
Reply

Tags
de novo assembly, ngs, short reads, velvet

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 02:35 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO