Error in filtering using DADA2

[thermokarst]

@hongwei2017

I have double checked my raw sequence data, it seems like the reads number in forward and reverse is not equal.

Yep, I have pointed that out now several times in this thread.

@hongwei2017

Is there anything I can do to make the both end sequences even before going for the denoising step?

There isn't anything in QIIME 2 for this, but you can try to run the following to help diagnose:

qiime tools export paired-end-demux-ITS-paired-end.qza --output-dir debugging
cd debugging
for f in *.fastq.gz; do r=$(( $(gunzip -c $f | wc -l | tr -d '[:space:]') / 4 )); echo $r $f; done

Theoretically you should see output that looks similar to this:

1046 YUN3428.1_30_L001_R1_001.fastq.gz
1046 YUN3428.1_30_L001_R2_001.fastq.gz
1338 YUN3428.2_45_L001_R1_001.fastq.gz
1338 YUN3428.2_45_L001_R2_001.fastq.gz
954 YUN3428.3_13_L001_R1_001.fastq.gz
954 YUN3428.3_13_L001_R2_001.fastq.gz
897 YUN3533.1.1_37_L001_R1_001.fastq.gz
897 YUN3533.1.1_37_L001_R2_001.fastq.gz
861 YUN3533.1.2_51_L001_R1_001.fastq.gz
861 YUN3533.1.2_51_L001_R2_001.fastq.gz

Every two lines should correspond to a single sample (forward and reverse), the two number should be identical (these are the number of raw reads in the file) - you are looking for the pair that isn't matched. Keep us posted!