Rapid PCR Barcoding data as methylation free?

[mtisza1]

Hi Alan,

I've returned with another question. I know that you say that Nanodisco requires ONT sequenced WGA DNA for the methylation-free dataset. However, would data generated from the Rapid PCR Barcoding kit (SQK-RPB004) serve the same function? I ask mainly because I'm lazy and the Rapid PCR Barcoding protocol seems easier for multiplexing. While we are on the topic, are there other methylation-free DNA that could be generated/used for Nanodisco.

Thanks for any insight you may have on this.

Mike

[touala]

Hi Mike,

This is not a library preparation protocol we used on the day to day basis and I've never used this type of data with nanodisco. I would expect to see more GC-related coverage bias than with the WGA approach but not necessarily. If this is the case, it would mean a potential power loss for some genomic regions but this might not be problematic depending on the amount of data generated. You would also lose the advantage of longer reads if some assembly/polishing is necessary. I would still recommend generating WGAs but if you decide to proceed with this solution, checking the genomic coverage along the genome is a good idea.

Methylation-free DNA for nanodisco analysis needs to match the native data. Considering that, there are not many other methods than the PCR-derived ones, besides generating RM systems KOs. Of course, the latter requires much more knowledge about the studied strain and more time investment.

Please let us know how it goes if you end up attempting it.

Alan

[mtisza1]

Alan,

Thank you for your thoughtful reply. I understand that there may be too many unknowns for you to give a concrete answer. If I decide to attempt the PCR approach, I'll let you know whether it was successful.

All the best,

Mike

[mtisza1]

Hi Alan,

I did want to briefly follow up on this idea. I did a comparison of several genomes with 1) PacBio sequencing/methylation calling 2) Nanopore sequencing of Native DNA VS WGA DNA + Nanodisco (Nanodisco WGA) and 3) Nanopore sequencing of Native DNA VS PCR DNA + Nanodisco (Nanodisco PCR).

I modified Nanopore's Rapid PCR Barcoding protocol slightly by using 7.5 ng starting genomic DNA and 7:30 extension step during the PCR. My N50 lengths for this protocol were always 3.3 kb - 3.5 kb with very normal distributions. I only did the analyses on data where genome coverage was > 75X.

• With both "Nanodisco WGA" and "Nanodisco PCR" I get all the same 6mA and 4mC motifs as PacBio and I get additional 5mC motifs, in some genomes. So, no difference between WGA and PCR in this regard.
• The read coverage across the genomes was actually more even with data from PCR DNA than the data from WGA DNA. The average GC content of the genomes I tested was between 40% and 50%.
• The Rapid PCR Barcoding protocol is much easier, faster, and more scalable than the WGA protocol.
• The flowcell output was high in with both library preps and comparable to rapid barcoding kit (i.e. native DNA).
• Although I did the T7 Endonuclease step in the WGA protocol, a couple of my preps still had some reads with low quality scores, reflecting incomplete digestion at this step. (This is probably a user error thing, but I thought I would mention it as a potential pitfall)
• I didn't compare the de novo assembly/polishing merits of the two datasets, but I imagine your point about WGA data being superior for this pipeline is correct.

I'm not sure if you or other researchers are interested in this protocol modification, but I am happy to have more formal discussions about it.

Mike

[touala]

Hi Mike,

Thank you very much for the analysis and report. This would be great if the Rapid PCR Barcoding protocol is as good as the WGA one. The more option the better. However, we would need to take a closer look at the data before mentioning it our guidelines and inform users about the potential benefits and drawbacks. Would it be possible to share some of your matched datasets in private (alan.tourancheau@bio.ens.psl.eu) so we can perform additional tests (e.g. methylation signal properties, and de novo assembly)? We engage to use them only for this analysis and promptly delete them. Of course, we will share the evaluation results with you as we go.

If you have questions, we can continue the discussion through email.

Alan