benjjneb
commented
4 years ago The key unanswered issue here is what is the amplicon you are sequencing? It is 16S, but what 16S region and what primer set? This determined how long the expected amplicon is, and affects what appropriate truncation length parameters are (they must be long enough to overlap).
Background: We have sequenced 10 samples of metagenomics data using illumina ampliseq. Its is 16S metagenomics panel. The data set consist of 301bp read length, paired end reads. The quality of the data is drops after 200bp.Therefore we decided to compare the reads with respect to Q20/Q25/Q30/full reads to trim the reads for merging(switching).As a result we got q25 data is merging better than all others. Sharing the parameters used(DADA2 script) as well as the resultant excel for your reference.
Key Point: Full raw data: Forward read length 301bp-Reveres read length 301bp Q20 Quality: Forward read Length 301bp- Reverse Read Length 260bp Q25 quality: Forward read length 285bp- Reverse read length 235bp Q30 quality: Forward read length 215bp- Reverse read length 190bp
Queries:
When we used the raw data(full data) without any trimming the data are getting filtered in huge number as you see in table. what could be reason for it?
Where seeing the comparison table we realized that Q25 quality (forward read length 285bp- reverse read length 235bp) seems to have optimal merged reads(highlighted bold). Can we consider that as optimal and take it forward for further analysis.
As per Dada2 software parameters there has to be over lap of 12bp minimum between forward and reverse reads. As we have trimmed the reads as per the quality still the merging happens in q25?! can you help us by explaining the possible reason behind this. I.e. which algorithm is been used behind for such kind of situations? 10_samples_DADA2_output_comparison.xlsx Example_dada2_scripts.txt