Alignment tool for long reads?

[naragam]

Hi,

I'm just starting to work with some long reads from a PacBio sequencer (>1Kbp) and I see that my usual alignment tools like MEGA, DNA STAR, bowtie, bwa are all restricted to smaller length bp seqs (<500 bp). Does anybody have good experience with alignment tools that can handle longer reads of say >1Kbp and upto 2.5 Kbp reads?

TiA, Nash

[honey]

long reads

MAY be BWASW

[SeekAnswers]

I used Blat to do reference based scaffolding by aligning contigs to scaffolds. So that should work with long reads I assume.

[honey]

Blat

Agree Blat is another good option

[pacbio]

Hi All

We have actually developed a fast and accurate aligner named BLASR (Basic Local Alignment with Successive Refinement) - http://www.smrtcommunity.com/SMRT-An...gorithms/BLASR to align our long reads. The source code for this as well as the full analysis software suite is freely available at the same PacBio DevNet site. A publication on this algorithm is also currently in review, so stay tuned.

[adaptivegenome]

Hi All

We have actually developed a fast and accurate aligner named BLASR (Basic Local Alignment with Successive Refinement) - http://www.smrtcommunity.com/SMRT-An...gorithms/BLASR to align our long reads. The source code for this as well as the full analysis software suite is freely available at the same PacBio DevNet site. A publication on this algorithm is also currently in review, so stay tuned.

Thanks for posting. I look forward to the publication as I am also trying to map PacBio reads.

[naragam]

Re: Alignment tool for long reads

Thank you all for suggesting blat and bwa sw for aligning long reads. I am looking at the docs for bwa sw and it looks like in the command:

bwa bwasw database.fasta long_read.fastq >aln.sam

the parameter "database.fasta" is the reference and the parameter "long_read.fastq" is the sequence being aligned. Right?

So does it absolutely need the fastq file or can it just work w/o the quality data, i.e., just a *.fasta file? Also how about the ccs based output from PacBio? Anybody has tried the PacBio ccs outputs? I'm trying to get "blasr" tool from PacBio pipeline installed here, but I am not there yet...

Thanks in advance,

Nash

[mchaisso]

Hi Nash,
You can install blasr on your own using github (https://github.com/PacificBiosciences/blasr).

If you have the hdf files, there are options (-useccsdenovo) to align the ccs sequences instead of the raw subreads.

HTH,
-mark

Thank you all for suggesting blat and bwa sw for aligning long reads. I am looking at the docs for bwa sw and it looks like in the command:

bwa bwasw database.fasta long_read.fastq >aln.sam

the parameter "database.fasta" is the reference and the parameter "long_read.fastq" is the sequence being aligned. Right?

So does it absolutely need the fastq file or can it just work w/o the quality data, i.e., just a *.fasta file? Also how about the ccs based output from PacBio? Anybody has tried the PacBio ccs outputs? I'm trying to get "blasr" tool from PacBio pipeline installed here, but I am not there yet...

Thanks in advance,

Nash

[naragam]

Thank you Mark...I don't have access to hd5 files yet....they are with the core sequencing facility and I am not sure they will give me those right now.... But I am working with them to gradually get some of the pipeline tools locally on my new Ubuntu machine that still needs memory upgrades before I can run your pipeline tools...

Yeah, I hope to run blasr soon but, in the meantime, I am trying to learn some of these long read tools that I haven't worked with before. Do you know if you have to have the fastq files for bwa sw?

Nash

[mchaisso]

Yeah, I hope to run blasr soon but, in the meantime, I am trying to learn some of these long read tools that I haven't worked with before. Do you know if you have to have the fastq files for bwa sw?

Nash

bwa sw aligns fasta sequences.

You will want the bas.h5 files since they have additional information about subread coordinates.

[naragam]

blasr compilation

Mark,

Am trying to compile blasr on my machine and am missing some header files in the tar file distribution. Can you please point me to sources who can help me or provide the *.h files needed? Thanks much,

Nash

[JackieBadger]

A quicker refined flavour of BLAT is BFAST :http://www.plosone.org/article/info:...l.pone.0007767

[naragam]

PacBio &quot;blasr&quot; questions....

Perhaps, I should really start a new thread...but, does anybody on this forum have good experience with blasr alignments to discuss the various options for the run and further the several output formats that are available. I have just started playing with some of the balsr runs and I have some pointed questions that I'd like to ask and/or seek detailed docs to refer to in terms of understanding all the options and outputs.

Any help available in this forum?

Thanks much in advance for any pointers,

Nash

[mchaisso]

Perhaps, I should really start a new thread...but, does anybody on this forum have good experience with blasr alignments to discuss the various options for the run and further the several output formats that are available. I have just started playing with some of the balsr runs and I have some pointed questions that I'd like to ask and/or seek detailed docs to refer to in terms of understanding all the options and outputs.

Any help available in this forum?

Thanks much in advance for any pointers,

Nash

You could say I'm pretty familiar with blasr output (I'm the author).

Most of the help may be found by running blasr -h, or blasr -help for detailed help. There are many output formats including tabular ones for which you can get column labels with the -header option, human readable output (-m 0), and sam (specified by -sam).

-mark

[naragam]

blasr output

Mark,

That's great to know....I have printed out the help pages, but there are still unanswered questions for me...would you like to take this discussion offline or do you want me to post the questions right here? If there's a special PacBio support site for blasr, I can reach you through that...Please let me know your convenience. Thanks much,

Nash