Assembling Sequences Without The Fastq File

Question

[Saad Khan]

I have been given a fasta file having short map reads obtained from some next generation sequencing method i have not been provided with the fastq file.What would be the best tool or best way to assemble these sequences either denovo or using a reference database. I was looking at bioperl howto : Short-read assemblies with BWA but it seems that it requires fastq files too. Please let me know the best way to go about in getting the best assembly possible.

[Ketil]

BWA maps reads to a reference, and you should be able to do that by just adding fake qualities, like Pierre suggests - possibly using decreasing quality towards the end of the reads. Many de novo assemblers ignore quality anyway (typically using de bruijn graph assembly), but I haven't been able to get very good results from them.

Answer 1

[Pierre Lindenbaum]

An idea: if you don't have the qualities of your reads(!), you could create a dummy fastq file with an average quality for all the bases and then use the standard software for assembling the genome ?

[Saad Khan]

So one way I can see it is that first i should make a Qual file with a numerical value of 50 for each nucleotide and then merge sequence and quality files to FASTQ using this code http://www.bioperl.org/wiki/Merging_separate_sequence_and_quality_files_to_FASTQ

Answer 2

[Michael Schubert]

Another idea: contact your supplier and ask for the fastq file.

If that's not an option, MIRA can do assemblies on fasta files. Velvet too, if I recall correctly.

[Michael Schubert]

Btw, I'm aware that BWA aligns to reference genomes and those are primarily de novo. But you haven't told us if you plan to align reads to reference, contigs to reference, or have a reference at all ;-)

[Saad Khan]

I do plan to use a reference genome. What if I do what Pierre suggested by making a Qual file with a numerical value of 50 for each nucleotide and then merge sequence and quality files to FASTQ using bioperl code (possibly decreasing quality at the end).Which one would be better using MIRA or using hypothetical quality

[Saad Khan]

Actually I am doing the assembly of a baculovirus, there are many genomes available at NCBI belonging to family baculoviridae. Are there specific assemblers for prokaryotic or small organisms, currently I am trying my hand with minimus which is a part of AMOS package. Please let me know what all I can do with the assembled genome.

[Michael Schubert]

http://en.wikipedia.org/wiki/Sequence_assembly#Available_assemblers

[Saad Khan]

Can you please provide me a link to a good tutorial on velvet, there is one btw in "Current Protocols in Bioinformatics" but I don't have access to it,If you have it please mail me at [e-mail]

[Michael Schubert]

Velvet has an excellent documentation on the website. Just look it up.