When I checked the output summary from DADA2, "percentage of input non-chimeric" came out to be low. (about 7 ~ 32% of input)
After removing the chimera, the absolute value of the read frequency seems sufficient because it's about 5K ~ 20K,
but the percentage of input seems too low, so I am not sure if there will be any problem such as data distortion in continuing the analysis
Please advise on what I should do.
I performed truncation based on Phred score 24.
This does seem quite low - have you removed the primers from your sequences using cutadapt? That would be my first recommendation (if you haven't already done so).
If running cutadapt on your sequences doesn't improve the high percentage of chimeras, then it may have something to do with your library preparation. How many PCR cycles did you use in the initial amplification step? How was the input material (good/high quality, high/low quantity)? That may have something to do with the high percentage of chimeric reads.
Here is an old forum post that discusses a similar issue that you may find useful for some additional troubleshooting: