Assembly paired-end sequences with Velvet

[efsalcedo]

I have sequences from illumina HiSeq 2000. I'll show you the sequence data that I have a bacteria.

fernando@InvDllo04[Lactobacillus] grep -c "^@" *.fq
FCHCTVLADXX_L1_wHAIPI016080-93_1.fq:1106031
FCHCTVLADXX_L1_wHAIPI016080-93_2.fq:1106031
FCHCTVLADXX_L2_wHAIPI016080-93_1.fq:1088418
FCHCTVLADXX_L2_wHAIPI016080-93_2.fq:1088418


As you can see, I have four files with 1106031, 1106031, 1088418, 1088418 sequences. I'm trying to perform an assembly with Velvet. but I am very confused because I do not know how to use these files. I was reading in some blogs and they say it is necessary to use a script from Velvet (shuffleSequences_fastq.pl), i.e.

1) shuffleSequences_fastq.pl FCHCTVLADXX_L1_wHAIPI016081-94_1.fq FCHCTVLADXX_L1_wHAIPI016081-94_2.fq FCHCTVLADXX_L1_wHAIPI016081-94_shuffled.fq

2) shuffleSequences_fastq.pl FCHCTVLADXX_L2_wHAIPI016081-94_1.fq FCHCTVLADXX_L2_wHAIPI016081-94_2.fq FCHCTVLADXX_L2_wHAIPI016081-94_shuffled.fq

3) velveth auto 31,45,2 -fastq -shortPaired1 FCHCTVLADXX_L1_wHAIPI016081-94_shuffled.fq -fastq -shortPaired2 FCHCTVLADXX_L2_wHAIPI016081-94_shuffled.fq

4) velvetg auto_31 -exp_cov auto

I do not know if I'm doing the right thing using the shuffled scritp or if I should just use the option "-separate" and run Velvet as follows:

velveth Assem 31 -shortPaired -fasta -separate left.fa right.fa

where:
-separate: Read 2 separate files for paired reads
-short -shortPaired
-short2 -shortPaired2

And the run "velvetg"...

My confusion here is because I have apparently two files "left.fa" and two "rigth.fa" instead of one as in the previous example.

I'd like you to tell me how to run Velvet

Thank you very much.

[bastianwur]

Don't have experience with velvet, because I found the setup very un-intuitive, and couldn't be bothered with it -> therefore some recommendation to you, use something else then.
e.g. Ray is easy to setup, and can handle more than one library.

[Brian Bushnell]

If you want to use velvet, I would concatenate your two read1 files together, and concatenate your read2 files together, then run it:

cat *_1.fq >r1.fq
cat *_2.fq >r2.fq

velveth Assem 31 -shortPaired -fastq -separate r1.fq r2.fq
velvetg Assem

A value higher than 31 is probably better for kmer length; that's just an example.

[Bioinform]

Next seq data with velvet

How to run velvet with Next seq data. It has four single reads??

Thanks
Bioinform

[Brian Bushnell]

Are you talking about 2 paired NextSeq libraries? Or do you mean Nextera LMP, which ends up (typically) with 4 output files for a single library after you process it?