UGENE Forum - Update and DNA Assemby

Nov 4^th, 2010 at 7:59am

Eitel Offline
YaBB Newbies

Posts: 5

bioinformatics gentlemen: I am new to using Ugene and I will make you a couple of questions: 1) I have the version 1.7.2 and have seen that there is at 1.8.0, how to upgrade?, 2) How do I make an assembly DNA that will allow more than the 3 mismatch that are by default? I work with sequences obtained by Sanger and I should do assemblies of about 400-600nt, and as is known, in an assembly of several readings is more than 3 mismatches!!!!. Consequently returns me an error saying:
[Tasks] [ERROR] Task} {DnaAssemblyMultiTask finished with error: DnaAssemblyToRefTask Subtask {} is failed: Bowtie TLS Task Subtask {} is failed: Reference assembly failed - not found possible alignment.
How do I make a DnaAssembly more "permissive".
I hope you could answer.

regards
Eitel

Back to top

IP Logged

Reply #1 - Nov 5^th, 2010 at 3:29pm

Agu Offline
Junior Member
(Argentinian)

Gender:
Posts: 63

Eitel:
I think that currently that's not possible. Your only choice with UGENE now is to manually assemble with the option "lock scales" (not funny if you have more than two readings...).

UGENE team:
I join to the request. With that feature I would be able to move almost completely to UGENE. Are you planning in implementing algorithms like CAP3 or phrap? Maybe it is possible to adapt UGENE's assembly tool (I have no idea).

Thanks

Back to top

IP Logged

Reply #2 - Nov 5^th, 2010 at 6:37pm

Mikhail Fursov Offline
YaBB Administrator

Gender:
Posts: 162

Folks, there a lot of holidays in Russia and this week is one such weeks

This means that the major part of UGENE team is far from internet and will be able to response only Nov-8 or latter

I could answer the questions, but not completely

1) Yes. Assembling & cloning is 1 priority for us for the nearest future. It means that we plan to
a. improve bowtie support,
b. optimize our own algorithm (why not to beat bowtie in? we have ambitions

)
c. We plan to add a special viewer to visualize assembly in UGENE. Now simple alignment viewer is used that is optimized to handle with (M)egs of data. But... assembly uses (G) prefix... so we will start new viewer development in 2-3 week.

2) As for BOWTIE question and 3 mismatches limit - lets wait for an expert who integrated it.

---
UGENE team

Back to top

IP Logged

Reply #3 - Nov 6^th, 2010 at 3:07am

Konstantin Okonechnikov Offline
Global Moderator

Posts: 173

Quote:

I have the version 1.7.2 and have seen that there is at 1.8.0, how to upgrade?

It's pretty simple: download the new version and install it. You can remove old version beforehand, but this is optional.

Quote:

How do I make an assembly DNA that will allow more than the 3 mismatch that are by default?

Actually Bowtie aligner is far more permissive then UGENE Genome Aligner. The -n mode (used by default) allows to set the number of mismatches in the so called "seed". According to Bowtie documentation in the -n mode the mismatches outside the seed are not taken into account while aligning. The seed is located at the left (high-quality) end of the read. The size of the seed is rather small (by default it is 28). You can make Bowtie even more permissive by decreasing the seedlen value.

Quote:

UGENE team:
I join to the request. With that feature I would be able to move almost completely to UGENE. Are you planning in implementing algorithms like CAP3 or phrap? Maybe it is possible to adapt UGENE's assembly tool (I have no idea).

As Mikhail mentioned we are planning to improve significantly DNA assembly functionality in UGENE. We will surely consider Pharp and CAP3 as candidates for including.

Back to top

IP Logged

Reply #4 - Nov 12^th, 2010 at 1:54am

Eitel Offline
YaBB Newbies

Posts: 5

Konstantin Okonechnikov wrote on Nov 6^th, 2010 at 3:07am:

Quote:

I have the version 1.7.2 and have seen that there is at 1.8.0, how to upgrade?

It's pretty simple: download the new version and install it. You can remove old version beforehand, but this is optional.

Quote:

How do I make an assembly DNA that will allow more than the 3 mismatch that are by default?

Actually Bowtie aligner is far more permissive then UGENE Genome Aligner. The -n mode (used by default) allows to set the number of mismatches in the so called "seed". According to Bowtie documentation in the -n mode the mismatches outside the seed are not taken into account while aligning. The seed is located at the left (high-quality) end of the read. The size of the seed is rather small (by default it is 28). You can make Bowtie even more permissive by decreasing the seedlen value.

Quote:

UGENE team:
I join to the request. With that feature I would be able to move almost completely to UGENE. Are you planning in implementing algorithms like CAP3 or phrap? Maybe it is possible to adapt UGENE's assembly tool (I have no idea).

As Mikhail mentioned we are planning to improve significantly DNA assembly functionality in UGENE. We will surely consider Pharp and CAP3 as candidates for including.

when I try to make Bowtie aligner with -n mode = 3, as suggested, and add 10 sequences in fasta UGENE suddenly closes!!! Any idea what could be the problem?
Thanks
Eitel

Back to top

IP Logged

Reply #5 - Nov 12^th, 2010 at 1:55am

Eitel Offline
YaBB Newbies

Posts: 5

when I try to make Bowtie aligner with -n mode = 3, as suggested, and add 10 sequences in fasta UGENE suddenly closes!!! Any idea what could be the problem?
Thanks
Eitel

Back to top

IP Logged

Reply #6 - Nov 12^th, 2010 at 2:50pm

Konstantin Okonechnikov Offline
Global Moderator

Posts: 173

Eitel, it looks like some nasty bug. More details are required to figure out the problem.

What UGENE version do you use? What is your operating system?

Did you use prebuilt index? Were the shortread sequences in separate files or in a single FASTA? What other Bowtie options did you use?

It would be great, if you could provide us with your data, so we can test it.

Back to top

IP Logged

Reply #7 - Nov 16^th, 2010 at 7:14pm

Eitel Offline
YaBB Newbies

Posts: 5

Konstantin

What UGENE version do you use? 1.7.2

What is your operating system?Ubuntu 10.04

Did you use prebuilt index? No idea what it is!!

Were the shortread sequences in separate files or in a single FASTA? shortread sequences in separate files
What other Bowtie options did you use? only that I mention

Regards
Eitel

Back to top

IP Logged

Reply #8 - Nov 23^rd, 2010 at 8:12pm

Eitel Offline
YaBB Newbies

Posts: 5

I have found another issue that causes UGENE to close unexpectedly and this is when I try to open more than 5 ABI files! it start to open a couple an then.... close!
It's not a memory problem because I have a lot (32Gb).
Thanks for your time

Eitel

Back to top

IP Logged

Reply #9 - Nov 25^th, 2010 at 6:54pm

Mikhail Fursov Offline
YaBB Administrator

Gender:
Posts: 162

Eitel wrote on Nov 23^rd, 2010 at 8:12pm:

I have found another issue that causes UGENE to close unexpectedly and this is when I try to open more than 5 ABI files! it start to open a couple an then.... close!
It's not a memory problem because I have a lot (32Gb).
Thanks for your time

Eitel

I can't reproduce this bug with ABI files I have.
Could you share the files that causes the crash?

You can use email address (ugene@unipro.ru) if you want to send them privately.

---
UGENE team

Back to top

IP Logged

	Welcome, Guest. Please Login or Register
	Welcome to our forum.