Quality Filtering of assembled reads central region

[svito]

Dear all,
I have ecountered some problems on my latest Miseq run with 2x300 kit (as I see from posts on the forum, is quite a common problem at the moment) and ended up with read1 and read2 with early dropping sequence quality.

I am analyzing the data now and first of all I used pear (with default settings) to join the reads, ending up with a per base quality like the one reported in image attached.

The question are:
1) first of all, should I care of the low-quality region in the middle of the assembled read? Or is it not really that bad?

2) the low quality bases at the end of the assembled are easily handled with a trimmer, I used sickle and worked fine, but how should I handle the low quality region in the middle of the assembled?


By the way, this is my first post, but I'm browsing the forum from some time now, and I want to take this occasion to thank you all for the precious information I got!!

[WhatsOEver]

Hi svito,
1) If I got it right, this is a plot of the merged reads, isn't it? In that case, what you are seeing is probably just an overlap of the quality drops towards the end of the individual forward/reverse reads. This is quite normal and your quality drop is not really that big. I would not care about it as long as your mapping rates are not drastically lower than expected.
2) Depending of what you're after, you could go for quality trimming before merging of the reads. But in your case I don't think quality trimming is necessary at all.

[GenoMax]

Most of your reads still appear to be above Q24-25 so that is not so bad. Depending on what you are doing I concur with WhatsOEver that trimming may not be needed.