Reads do not align with reference genome

[chavesbio]

Hello community.

I am new to NGS world and I'm facing some problems while trying to align my reads with a reference genome.

First, let me explain what I'm doing. I am working with a plasmid of about 44 kb that I've sequenced with Illumina MiSeq. Now, I sequenced 7 other plasmids from the same species (different strains). I used an Ion PGM for these other 7 samples.

Now, I was using bowtie2 to align my reads with the plasmid's genome that I've sequenced with the Illumina's platform and almost all the reads (from 260,000 to 360,00 depending on the sample) do not align with my reference genome.

Do you think I might be doing something wrong in this process? Or there's the possibility that I have something wrong with the sequencing (the contigs generated with the assembler plugin from Ion's platform do not align with the plasmids)?

Thank you in advance.

[GenoMax]

Things could be wrong at either end e.g. MiSeq plasmid not assembled right or ion assemblies are not good. Sounds like you expect the plasmids from the 7 other strains to be similar to the one you did with MiSeq.

How did you assemble the sequence for the "MiSeq" plasmid? Did you try getting the fastq files from ion to see if they align with bowtie2?

I would suggest trying a non-NGS aligner (e.g. BLAT from UCSC) with the ion contigs to see if you are able to align things. MAUVE also would be a great option since you are working with small plasmids.

[chavesbio]

Hello, GenoMax.

That's right. I expect those other 7 plasmids to be similar to the first one I've sequenced. To sequence this first one, I payed an enterprise to do the job for me. To help them in the assembly, I sent them some contigs from this plasmid that I obtained from Sanger seq. They used Velvet for the assembly. At the end, the result was satisfatory.

I got the Fastq files generated from Ion (apparently, they're already precessed with the adaptors removed and the quality cutoff).
When using bowtie2, first I got the complete sequence of my first plasmid and created the index files from it. Then, I gave a command line to align the reads of one newly sequenced plasmid with the reference genome. At the end, the software gave me the numbers of sequences aligned and the result was very bad (0% of the reads were aligned with the reference genome)

I'll try to BLAT and MAUVE.

Thank you for your attention.

[GenoMax]

That seems odd. If the MiSeq plasmid was correctly assembled and if next 7 plasmids are similar to the first one then you should have seen some alignment as opposed to none.

Let us know what blat/mauve do (you will need to convert your fastq files to fasta for use with blat/mauve).