Hiseq2000-Quality filtering

[bioman1]

I am newbie in NGS. We have sequenced non-model plant through illumina Hiseq 2000, and I preprocessed the raw sequence with fastqc, with fastqc quality check, the sequences passed all the test but failed in "Per Base Sequence Content". What should I do?. In what way I can improve the per base sequence content by quality filtering?. What type of quality filtering/tool should I use?

[GenoMax]

Can you post an example plot? If one of the items "fails" in FastQC then it does not automatically mean a strike against the dataset.

Here is a recent post of mine with a list of trimming programs/adapter information: http://seqanswers.com/forums/showthread.php?t=40054

[bioman1]

re:fastqc

Thanks for the reply. Below is the fastqc graph in which per base sequence content and duplication level gets warning in fastqc. Is there need to filter?

[GenoMax]

The question you should be checking into is if there is adapter contamination and/or low quality sequence that you need to remove before the downstream analysis. Some links for various trimming programs are in: http://seqanswers.com/forums/showthread.php?t=40054

Also check post #2 in http://seqanswers.com/forums/showthread.php?t=39889 for an insight into the sequence duplication "problem".

What kind of an experiment is this? It is normal to see that kind of per sequence base content for illumina sequencing.

[flobpf]

Is this RNA-seq or genomic sequencing? Can you post a few more details about the expt? What are you going to do afterwards? Assembly?

From what you've posted above, the %AT/%GC looks too uniform (this is how it should be: http://www.bioinformatics.babraham.a...qc_report.html), however, it may not necessarily be an issue.

I don't think there is a trimming issue here, since the ends look fine. You might want to go ahead with the assembly as planned and BLAST the assembled sequences against the NR database in NCBI, just to make sure that they are mapping to sequences close to your species.