Questions on filter and trim removal of reads

[cdesouza02]

Hi all,

I am still quite new to dada2 and have some weird issues when running the Filter and Trim function of dada2.

I am running dada2 on Illumina MiSeq300 paired-end sequences and I trimmed off the primers (515f-806r) and adapters with cutadapt prior to running dada2. But when I applied truncLen=c(250, 225) and maxEE=c(2,2) which I read off from my plotQualityProfile plots, corresponding to the quality drop, 96% of my reads were filtered off.

reads.in=190092
reads.out=7276

Before dada2, I trimmed the 16S adapters off using cutadapt. In dada2, I am using trimLeft because I saw (using fastqc) that my samples have odd (non-adapter) sequences from 0-10 bp.

Do you have any advice or any different methods I could try to retain more of my reads?

Thanks!

[cdesouza02]

I had a higher number of reads kept when I used untrimmed sequences and TrimLeft to include the primers.
However I have N-base pairs (that are now A,G,C, or T) before my primers. Samples vary in number of N's (i.e. 0N, 1N, 2N and 3N) before each primer and I also want those trimmed off.

These N's became a problem in my downstream analysis with phyloseq, as the samples are clustering by number of N's before the primer instead of a real variable.

Are there any recommendation for trimming in dada2 where I could do all of the samples in one command (to avoid biases), while also taking the N's out?