100% reads passed chastity filtering?

[savernake]

Hi, folks.

I'm on my fist QC analyses and when filtering for failed chastity reads the program retuned that 100% of the reads passed the test (used fastq_illumina_filter). I did not work on the sequencing (mammal on Illumina HiSeq, paired end), although i'm almost sure it was not processed before it fell on my lap. Could it be possible that that is indeed a true result? Or am i doing something wrong?

"Processed 12,162,651 reads
fastq_illumina_filter (--keep N) statistics:
Input: 12,162,651 reads
Output: 12,162,651 reads (100%)
Processed 12,162,651 reads
fastq_illumina_filter (--keep N) statistics:
Input: 12,162,651 reads
Output: 12,162,651 reads (100%)"

Thanks in advance for the help.

[mastal]

I'm not familiar with fastq_Illumina_filter, but I think that only the reads that pass Illumina's filters get output as fastq files to send to the end users, so unless you are actually a service provider working with the sequencer, you probably wouldn't see any reads that don't pass the purity and chastity filters.

If the reads haven't been processed further, then you can expect them tobe all the same length, and some will have N bases or adapter sequences. You can use FastQC to look at base qualities, etc.

[GenoMax]

Use FastQC for initial QC of your data. It is much more intuitive. Don't be afraid of red "x" showing up on the reports since they may be related to type of experiment you are doing.

Take a look at this site for commentaries on possible results from FastQC.

[savernake]

Thank you for your answers!

I indeed did a FastQC analysis that returned some fair results. I was trying to filter the reads so I could improve the base quality scores (image below) following the recommended by Zhou and Rokas 2014. But as it seems the results could be really correct I guess i'll pass to the next level and start trimming the bad reads/3' tail.

[cpad0112]

I can't comment on 100% quality reads, but as you mentioned reads need 3' end trimming. Please post your qc plots after trimming, if you do not mind.

[savernake]

I can't comment on 100% quality reads, but as you mentioned reads need 3' end trimming. Please post your qc plots after trimming, if you do not mind.

Hi!

I finally managed to trim my sequences. The per base quality increased as expected, although I'm still getting some strange pattern on k-mer content. I'll post the results bellow (in order: original, adaper trim and quality trim for kmer-content and adapter trim and quality trim for base quality

[GenoMax]

Generally there is no need to worry about k-mer composition. If you are happy with the trimming results move forward with the rest of the analysis.