Webinar: Nanopore Sequencing in Plants: From Greenhouse to Genome
Answers from Maximilian to your questions during the webinar:
1. Sample prep - what is succeeding for people? Are nuclear isolations necessary, or do people succeed with CTAB and column based enrichment of large fragments?What is meant by ultra-long read? Is this just skimming off long reads from a regular run or different? Any recommendations about premature clogging of pores? Even after clearing and following the protocol for pore refresh, it clogs again quickly. Got 6-7 M reads not 14.Suggestions for lower per base error rate for genome sequencing?
I can't talk about what works for other people but for us most of the time non-nuclei direct CTAB methods are easier and work as well. However there are some species (e.g. Quinoa) that seem to have some pretty stable interaction with DNA if the get the chance to interact so your best chance there is to remove those by nuclei washes and only release the DNA afterwards.
I do not know if there is a definition for "ultra-long reads" I would say that means reads >100 kbp. And yes you do get several of those from normal runs. If you want to increase the fraction of those in a run you can either go for the Tagmentation based rapid Kit approach by Josh quick (https://www.protocols.io/view/ultra-long-read-sequencing-protocol-for-rad004-mrxc57n) or try to reduce the number of bead Clean-Ups in the normale LSK-109 protocol by either doing no clean-up between a-tail and ligation or precipitate with high salt and spin down instead of bead-clean up. However both will lower the yield from the run.
6-7 Mio reads for plant DNA sounds pretty normal for me for >25kbp N50 runs we also do not get more. You could try doing multiple nuclease flushes and reload over the course of the run with the Wash Kit version 3.
If you suspect the lower quality comes from DNA damage you could try if doing a different DNA extraction will help. Appart from that I think you can only wait since Nanopore is working on plant specific basecallers that should help there.
2. Which of the extraction methods uses the least harsh chemicals?
Probably the Circulomics Nanobind protocol but even there you will need 2-Mercaptoethanol for nuclease inactivation during nuclei extraction I think.
3. Would endopolyploidity be an issue?
I assume not since there should not much recombination happen so you just sequence more of the same chromosomes, but I never tried.
4. What would be the rough pricing for BioNano DNA extraction?
I do not know since we just asked to get the leftover DNA sent to us from the service center that did the optical mapping for us. I assume it would be below 200 € but you need a rotor-stator homogenizer which is what we are currently lacking to do it on our own.
5. Are the slides available anywhere?
The pdf of the lecture can be found here.
6. How would you determine the purity of DNA? For example, the contamination from polysaccharide or protein.
Proteins you can often see on the Nanodrop by a low A260/A280 I think that is pretty much the standard method, the problem there is that it only detects aromatic amino acids so some proteins might go undetected. Polysaccharides are even harder. We never found an assay sensitive enough to detect contamination in DNA without being affected by DNA itself. What works best for me is actually looking at the DNA at the end of the prep when you add the ethanol or Isopropanol to precipitate it. If you see defined structures like a clearly white defined structure with white strings that is overall clearly white then there are most likely not many polysaccharides in it. If it looks pretty transparent and more like a big “blob” when you move it out of the solution with a pipet tip or glas hook and it slides easily of those than there are probably a lot of polysaccharides in it. Sorry for being that vague but I actually do not have a better answer or method.
7. Do you and if so how often reload the flow cell? Do you Nuclease wash each time?
Yes I do reload MinION and PromethION cells, usually once per run but if I get the impression that there are still available pores at the end of the run I also do a second flush sometimes. I did use the nuclease flush protocol previously, but since the Flow Cell Wash Kit version 3 got released we use that once since it includes a nuclease flush.
8. What is your average output (bases and reads) per flow cell (MinION)?
That unfortunately widely varies between species and is also dependent on read length but on MinION I would say the yield is normally between 8 and 15 Gbp with about 300.000 to 600.000 reads. For Promethion usually 25 to 60 Gbp with 2.5 to 6 million reads.
9. BUSCO completeness was calculated with Eukaryota odb or Viridiplantae odb?
For Vinca I was using the ODB10 Eudicot set and for Physalis the ODB10 Solanaceae set
10. Thank you for the excellent lecture. Is the presentation going to be available for us after this lecture?
Thank you very much. The pdf of the lecture is in the same folder as this document.
11. What is your advice on Onion genome sequencing which is 16GB?
That is not going to be easy. I would start with 50x coverage (800 Gb) and see if wtdbg2 and your servers can handle it. It probably won’t be a great assembly and it will run several weeks I suppose but it should at least give you an idea how much more data you might need. But take my suggestion with caution since I have never worked with something that huge.
Maybe have a look at what the people who did the giant redwood genome used. I think they used Masurca with a combination of Nanopore and Illumina. They gave a talk at Nanopore’s New York conference last year
12. Are there any applications in your work for short read sequencing?
For denovo genomes we still use PCR-free short read data polishing and do Hi-C with short reads. We also still have several projects focusing on SNP’s for which we use short reads and everything related to differential expression we still do with short-read RNA-seq. But since I have no direct access to an Illumina sequencer anymore but prefer to do things in-house when I can I try to avoid it where I can.
13. Which extraction method would you advise for Arabidopsis thaliana?
I think I would go for the usual Qiagen column DNA extraction with Carlson buffers. You can find it on the Nanopore homepage (https://community.nanoporetech.com/extraction_methods). I have not tried it on Arabidopsis but it did work well on brassica napus. If you take all healthy looking leafs from one plant that should be enough else try growing a lot of seedlings in parallel and use the aerial parts.
Michael et al. 2018 (https://www.nature.com/articles/s41467-018-03016-2#Sec9) did basically use this but without the columns.
If it needs to be fast and easy I think in Arabidopsis also the Sorbitol CTAB method by Souza et al. 2012 (http://www.funpecrp.com.br/gmr/year2012/vol11-1/pdf/gmr1734.pdf) should work since its not heavy in secondary metabolites.
Very recently Jiao et al. 2020 (https://www.nature.com/articles/s41467-020-14779-y) did use the NucleoSpin® Plant II Maxi Kit from Macherey-Nagel. I have never used it but it did work for them when doing long-read sequencing
14. Can you recommend (teaching) literature on the bioinformatics analytics you were presenting today?
That is actually a good question. Unfortunately I don’t know anything from the top of my head. I was looking for this myself when I got into the topic but could not find much so I did the “learning by doing” method.
15. In RNA doesn't artificial light affect the expression of genes?
Yes it definitely does, when we are looking for differentially expressed genes we try to use as much natural light as possible since many genes are involved in the photo-cycle. However in what I presented in the webinar I was just talking about finding genes at all and not comparing expression rates so it was not so much of an issue there.
16. Can you please provide the slides for the presentation
The pdf of the lecture is in the same folder as this document.
17. What about plastidial SNA sequencing
Plastidial DNA I suppose? I was never really interested in plastid genomes especially so I did not check, but at least in short-read sequencing and assembly one was always getting the chloroplast genome for free even when using nuclei enrichment. I have no reason to believe it would be different in Nanopore assemblies. I suppose it will be one contig in an assembly.
18. What other ways do you recommend to check molecular weight of DNA and RNA without FemtoPulse or PFGE?
Before we had a PFGE we used to do 0,6% Agarose gels at 100 V for about 60 minutes with the DNA and a Thermo Fisher High range ladder with bands from 15 to 48 kbp. HMW DNA will still be pretty much a single band but the less of a smear downwards the more intact it is. Try doing some dilutions on the same gels since often DNA will look fragmented just because the gel lane is a bit overloaded.
For RNA I usually do native 1% Agarose gels at 100V for about 45 minutes. If I see two strong bands where the upper one is stronger and no smear below 200 bp in comparison to a DNA ladder running beside I assume that the RNA is intact. Admittedly this is not scientifically correct since you should be doing denaturing gels and us a RNA not a DNA ladder, but it does work for me so far. Optimally I think you should be doing Bioanalyzer or Tapestation analysis since they will provide you with RIN (RNA integrity number) which is the generally accepted way of assessing RNA quality but I do not have direct access to one of those.
19. Is it possible to do differential gene expression analysis using transcriptome data sequenced using Nanopore sequencing?
I would say yes but I have never tried it and see some problems there. In theory you will get enough reads from a MinION to be enough to compare to another sample running on a different MinION but since we still see varying yield between different cells with similar samples I do not know how this would influence the result. On a PromethION it should also be possible to run several libraries with barcodes in parallel but I have heard that often enough barcodes get miscalled (maybe this is fixed by now) and that might lead to reads being assigned to the wrong sample. In data analysis I am also not sure how it would work since the normal normalization with transcripts per million reads or FPKM will probably not work with Long-Reads and the k-mer counter tools like salmon or kallisto won’t work either so I would say in theory yes but the bioinformatics part might be a bit more challenging than usual.
20. I'm having difficulties to extract RNA from grape leafs infected with a specific fungus using commercial kits. How was your experience using alternative protocols or kits?
Yes I know the pain. An easy approach you could try if you haven’t already is the Directzol kit from Zymo. It did work well for me most of the time but with just 1-2 µg yield from difficult tissue, but maybe if you do several in parallel it might work. If that fails Try Chang et al. method for pine needles (https://www.researchgate.net/publication/226940488_A_Simple_and_Efficient_Method_for_Isolating_RNA_from_Pine_Trees) You can downscale it to about 200 mg tissue and 1 mL extraction buffer in the beginning. It’s a bit more work than your usual kit but it worked very well for me. Alternatively you can also employ the Qiagen genomics tips. There is a protocol how to do this on the Qiagen homepage (https://www.qiagen.com/us/resources/resourcedetail?id=4bc22f81-e1ef-4271-a115-f215f7e9778f&lang=en). Just be careful to prepare the buffers nuclease free since you cannot DEPC-treat those. And rinse the pH electrodes with 100 mM NaOH before adjusting pH anywhere. I do get up to 90 µg RNA from ~600 mg root tissue this way.
I sorted them from lowest to highest effort. I personally would probably recommend the pine-needle most.
21. What is the best Sequencing way to determine insertion number and location of insert for transgenic plant
I have to admit that that is a problem I was never confronted with and have not thought about it.
22. Please send us this lecture via mail
The pdf of the lecture is in the same folder as this document
If you'd like any further information about plant sequencing with nanopore technology then contact our technical team at support@nanoporetech.com.