Since I visited qiime2 forum, my analysis process has been rocketed! Thanks for all people helped me!
Here I find an issue when I use cutadapt to demux my paired-end sequences. I checked my sequences, in forward sequences, all my samples were mixed in one file, and different barcodes are at the front of each read. And in reverse sequences, no barcodes can be found. But in the middle of some reads, I also find the sequence same as barcode. For example, there have two reads belong to two samples, sample one barcode: CTATAC. smaple two barcode: GTCCCA.
The first read: CTATAC.....................
The second read: GTCCCA................CTATAC........
So, when I use the cutadapt to get the demultiplexed seqs, is there have any possible that both these two reads will be recognized to sample one?
I did face this problem. Here is my result after demux:
The sample C1 gets impossible 300554 reads, when I tracked back to raw forward seqs file, I found in the middle of so many reads, there have some successive sequence same as C1's barcode.
Is there any way to fixed this problem?
Then I have another small question. I know it is easy for cutadapt to trim the primer seqs. And the sequencing company told me there have an anchoring seqs to connect the primer in 3' end. I am not sure the name anchoring is right or not. Anyway, this seqs: TGGAATTCTCGGGTGCCAAGGAACTC did in forward seqs, and I was puzzled to find it appeals in random location of some reads, but some reads have zero anchoring seqs. If I leave it in trimmed-seqs after cutadapt, and do the dada2 denoise directly, will it cause problem in final result? Or is there any way I can cut this anchoring seqs in some reads?
