Hello everyone,
What happens if I try to sequence with Iron torrent amplicons (eg of a marker gene) that are too long? By that I mean more than 500bp. We did try (not on purpose, we did not realised about the max length of around 430bp), and got a very random pattern of read length coming back. So do you know if the fact that the sequencing does not reach the end of the DNA strands creates a mess in the process?
Thanks for your help.
Regards,
Sven
The fragment library preparation kit is designed for amplicons up to 400bp. This length includes the primers and adaptors/barcodes. If your amplicon size is more than that, any results you might obtain from the sequencing will not be reliable. The phred score (Q20) might be very low which might not be sufficient for downstream bioinformatic analysis. You could either reduce your amplicon size by redesigning the primers or you could also shear the amplicon using restriction endonucleases.
Hei Nishanth,
thanks for the answer. But I think I formulated badly my question: Yes, of course it makes sense that the phred score will go down. However the phred score is assigned per bp, and therefore I would expect to have high phred score until ca 400 bp and then the quality would crash down. In my pipeline, I could then trim the reads to keep the 400 first bp, and remove the end of the sequence.
However, it seems that something goes into the way of the sequencing process as the sequences that came back were of many different lengths (see picture attached, tags 56 to 69). This "something" is what I do not understand.
Hi Sven,
It seems like you are using the 400bp library prep kit. From my understanding, the Torrent Suite already truncates the barcodes and primers while calculating the mean read length. The Q20 reads seem to be fine after the deletion of erroneous basepairs. I haven't done many runs on the Ion Torrent systems (maybe 6-10 runs) but even I have seen a range of read lengths across all my samples. Even the Thermo Fisher technician seemed to think that was normal. I feel the range you have obtained (177bp-272bp) is wide enough to doubt the efficacy of the sequencing run. Although my knowledge might be limited in this area, what I would have done would be to perform the downstream bioinformatics and then see if the data is relatable. If the barcoded samples are in duplicates, then you might be able to check how relatable the results are between the duplicates.
- Badges
- Science method