Velvet and reads from different generations of Illumina

[hiddenrisk]

Greetings colleagues!

I am a second year Ph.D. student and I am currently taking a genome assembly class. We have been given a class project to de novo assemble a set of reads. The data set that we have been given contains mate pair reads from two generations of Illumina sequencing- some of the reads are 36bp and some of the reads are 76 bp. The professor of the class suggested that Velvet may produce different assemblies if I separate out the reads by size and do my assembly with the reads of different lengths in different files. However, as these sequences aren't actually long (capillary) reads, I can't figure out how to run velvet with all of that information.

The velveth command I think I would run (I'm doing this in Linux) is...

velveth [output directory] [kmer] -short -fasta short_forward.fa short_reverse.fa long_forward.fa long_reverse.fa -shortPaired short_shuffled.fa long_shuffled.fa -singletons short_singletons.fa long_singletons.fa

Is this correct? I haven't tried it yet, mind you, but I cannot conceptually understand how or if velvet would be able to use this information properly.

I have done the assembly with the data contained in single files successfully, but was hoping to generate something better if it's possible.

Thank you for considering my question!

[hiddenrisk]

So, I stumbled into the velvet page on the seqwiki page and it answered my question, I think. I'll try doing what is suggested on the page, and let you all know how it works.

[hiddenrisk]

So, I managed to only sort of be successful... I cannot include any singleton data when I run my assemblies.

Does anyone know how I might do this?

[Ole]

From reading the manual, I think this should work:
velveth [output directory] [kmer] -shortPaired -fasta short_forward.fa short_reverse.fa -shortPaired2 long_forward.fa long_reverse.fa -short short_shuffled.fa long_shuffled.fa short_singletons.fa long_singletons.fa

It depends on what kind of insert size you have of course. Use -shortPaired for the two files of one insert size, and -shortPaired2 if you have a second set of pair ends files with a different insert size. Just use -short and list the singleton files after the option, like "-short short_singletons.fa long_singletons.fa".

Btw, the manual is at http://www.ebi.ac.uk/~zerbino/velvet/, and I think it will answer most of your questions. Hope this helps.

[hiddenrisk]

Thank you for your help! I read the manual, and also the wiki, and what I ended up doing was...

-short fasta short_fwd.fa short_rev.fa \
-shortpaired short_shuffled.fa \
-short short_singletons.fa \
-short2 long_fwd.fa long_rev.fa \
-shortpaired2 long_shuffled.fa \
-short2 long_singletons.fa ;

And it worked just fine.