Using dbcAmplicons Low pairing of reads from 16s rRNA data based on Illumina

Question

[olooo.odhiambo]

I am using dbcAmplicons pipeline for my data analysis of 16s rRNA from microbial communities in chicken gut based on Illumina sequence. The logic is to use paired reads yet only about 2 % of my reads are paired, the rest are not. What could be problem? and what is the implication of working with this too few no of paired reads down stream? PS: I am just learning the basics of command language and hoping my question makes sense? i appreciate your help.

Answer 1

[Tm]

Logically, when the illumina platform is used for 16s rRNA analysis, then paired end libraries are prepared and sequenced. So all sequences should be paired only. However, paired-end reads are first stiched/joined and then used for taxonomy assignment. I have not used dbcAmplicons pipeline, but I think in your case, you are ending up with 2% stiched reads which is extremely less and it can be a result of presence of adapter sequences in your paired-end reads.

[olooo.odhiambo]

Thanks @toralmanvar for your response. I realized adapter sequences are only about 8bp yet my average library sizes is 500bp. The reads on Illumina MisSq was 2x151bp I realized that this may result in gaps when joining the paired end reads and may result in low joined read%?

[Tm]

Can you tell us which variable region you have sequenced? Because if amplicon size is more than 300 bp, and you have sequenced it on 2 x 150 bp chemistry, then they don't have any overlaping region to join. Thus, preferably 16s amplicons (specially v3-v4 region) are sequenced on 2x250 or 2x 300bp chemistry.

Answer 2

[olooo.odhiambo]

Thanks gain @toralmanvar, I did sequence the v3v4 region of the 16s. Our team recommended the 2x151 MiSeq sequencing on the basis that their experience with such work is that when they sequence longer reads, 2x300 then the QC parameters of the data dips significantly. But I truly see your logic here!.