1) The document discusses best practices for maximizing throughput when using Nanopore technology, including ensuring high purity and integrity of input DNA samples. 2) It describes using Nanopore sequencing to generate de novo reference genomes for genetic lines of Arabidopsis thaliana, with high quality assemblies obtained. 3) Generating long reads with Nanopore allows detection of structural variations between genomes, with the potential to improve genome-wide association mapping.

1. @NanoporeConf | #NanoporeConf
Nanopore Day Paris 2018
BEST PRACTICE TO MAXIMIZE THROUGHPUT
WITH NANOPORE TECHNOLOGY
&
DE NOVO SEQUENCING OF GENETIC LINES OF
ARABIDOPSIS THALIANA.
Baptiste Mayjonade
French National Institute for Agricultural Research

2. BEST PRACTICE TO MAXIMIZE THROUGHPUT
WITH NANOPORE TECHNOLOGY
01

3. 3 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
WORKFLOW FOR 1D LIBRARY WITH GENOMIC DNA
DNA QC : Purity and integrity
DNA shearing
Adapter ligation
Flow cell loading
DNA repair + End-Prep
LIBRARY
PREPARATION
BluePippin step

4. 4 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
DNA QC: PURITY OF THE SAMPLE IS THE KEY PARAMETER
 Library prep : impurities can reduce the efficiency of the enzymatic reactions
 Sequencing : pores could be very sensitive to contaminants
 Ultrapure DNA is required to obtain optimal results
• Nanodrop not accurate for concentrations < 20 ng/µl
• A260/A230 = 2 - 2.2 (could be tricky to reach)
• A260/A280 = 1.8 - 2 (if >2 RNA contamination)
NANODROP QUBIT
Nanodrop is accurate for pure DNA
 Nanodrop concentration ≈ Qubit
Fluorescence dsDNA specific

5. 5 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
DNA QC: PURITY OF THE SAMPLE IS THE KEY PARAMETER
Equivalent purity but different behavior during the sequencing
A260/A280 = 1.93
A260/A230 = 2.05
Plant gDNA (20 kb sheared)
250 pores working (strand + single)  4-5 Gb
After 5 minutes of sequencing (mux1 = 504 active pores)
Even if DNA purity looks good nothing is guaranteed.
A260/A280 = 1.86
A260/A230 = 2.12
Bacteria gDNA (20 kb sheared)
471 pores working (strand + single)  10-15 Gb
After 5 minutes of sequencing (mux1 = 505 active pores)

6. 6 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
DNA QC: INTEGRITY OF THE INPUT DNA & ASSESSMENT OF THE LIBRARY SIZE
 Input material (gDNA) :
• if degraded  size selection to get rid of short fragments
• if no high molecular weight DNA  no long reads
 Library : accurate assessment of the library mean size  convert the DNA mass to a number of
molecules  using the appropriate amount of DNA as input for the library preparation
gDNA sample mean size = 10kb gDNA sample mean size = 50kb
0.2 pmoles = 1.32µg 0.2 pmoles = 6.6µg
Useful link: http://www.promega.com/a/apps/biomath/index.html?calc=ugpmols

7. 7 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
DNA QC: INTEGRITY OF THE DNA & ASSESSMENT OF THE LIBRARY SIZE
Comparison of 5 devices to assess the profile of a genomic DNA
FRAGMENT ANALYSER
FEMTO PULSE
PIPPIN PULSECHEF MAPPER
10 kb 20 kb 50 kb 100 kb
Average size = 53 000 bp
10 kb
Smear = 20 – 100 kbSmear = 20 – 150 kb 50 kb 150 kb
20 kb
100 kb
50 kb
Average size = 83 000 bp
48.5 kb
15 kb
97 kb
145.5 kb
100 kb20 kb
12 kb
Electrophoresis
Band > 12 kb

8. 8 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
DNA QC: INTEGRITY OF THE DNA & ASSESSMENT OF THE LIBRARY SIZE
Device
Electrophoresis
category
Run
time
DNA
requirement
Suitable Price
CLASSIC ELECTROPHORESIS 1h 100ng 0 – 12 kb ≈ 1 k€
CHEF MAPPER (BIO-RAD) Pulsed-field 16h ≈ 500 ng 0 – 10 Mb ≈ 30 k€
FEMTO PULSE (AATI)
Pulsed-field
capillary
2 - 3h ≈ 5 fg 0 – 150 kb ≈ 80 k€
PIPPIN PULSE (SAGE SCIENCE) Inverted 16h ≈ 100 ng 0 – 100 kb ≈ 5 k€
FRAGMENT ANALYSER (AATI) Capillary 1h ≈ 4 ng 0 – 50 kb ≈ 35 k€
FEMTO PULSE appears as the best tradeoff between analysis time, DNA requirement and
resolution for large DNA fragments

9. 9 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
DNA EXTRACTION KITS AND PROTOCOLS
Spin column
QIAGEN DNesasy Kit
12 kb
5 kb
ADNg Ladder
Magnetic beads
QIAGEN MagAttract
Flow gravity column
QIAGEN Genomic-tip
Sunflower Human
cells
E.coli
https://www.protocols.io/groups/minion-user-group-with-fungi-and-plants-on-their-mind
Other protocols available :
(Benjamin Schwessinger)

10. 10 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
WORKFLOW FOR 1D LIBRARY WITH GENOMIC DNA
DNA QC : Purity and integrity
DNA shearing
Adapter ligation
Flow cell loading
DNA repair + End-Prep
LIBRARY
PREPARATION
BluePippin step

11. 11 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
LIBRARY PREPARATION : DNA SHEARING
Comparison of 2 devices
G-tube
(1 pass through the g-tube)
Megaruptor
(15 passes through hydropore)
Wide distribution
30€ per sample
Only a centrifuge is needed
Tight distribution
10€ per sample
Megaruptor device ≈ 18k$
Tight distribution = less long fragments  more efficient library preparation (?)
 Upcoming new ligation kit (LSK109) more efficient for long fragments

12. 12 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
WORKFLOW FOR 1D LIBRARY WITH GENOMIC DNA
DNA QC : Purity and integrity
DNA shearing
Adapter ligation
Flow cell loading
DNA repair + End-Prep
LIBRARY
PREPARATION
BluePippin step

13. 13 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
LIBRARY PREPARATION : SIZE SELECTION (BLUEPIPPIN - SAGE SCIENCE)
To purify a sample : plant sample (A. thaliana)
A260/A280 = 1.88
A260/A230 = 1.98
A260/A280 = 1.86
A260/A230 = 2.29
With a BluePippin step (6 Gb)Without a BluePippin step (4 Gb)
337 pores working (strand + single)
After 5 minutes of sequencing (mux1 = 505 active pores)
143 pores working (strand + single)
After 5 minutes of sequencing (mux1 = 497 active pores)
A BluePippin step improves the purity of a DNA sample and has a positive impact on sequencing

14. 14 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
WORKFLOW FOR 1D LIBRARY WITH GENOMIC DNA
DNA QC : Purity and integrity
DNA shearing
Adapter ligation
Flow cell loading
DNA repair + End-Prep
LIBRARY
PREPARATION
BluePippin step

15. 15 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
LIBRARY PREPARATION : ADAPTER LIGATION
Overloading the ligation reaction could reduce the amount of DNA molecules with adapters
on both ends  ≈ 0.2 pmol into the ligation reaction appears optimal (8 kb and 20 kb library)
 Higher adapter concentration in the upcoming LSK109 kit (new ligation kit)

16. 16 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
WORKFLOW FOR 1D LIBRARY WITH GENOMIC DNA
DNA QC : Purity and integrity
DNA shearing
Adapter ligation
Flow cell loading
DNA repair + End-Prep
LIBRARY
PREPARATION
BluePippin step

17. 17 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
LIBRARY PREPARATION : FLOW CELL LOADING
45 fmols of a 20 kb library loaded onto the FC
After 5 minutes of sequencing
90 fmols of a 20 kb library loaded onto the FC
After 5 minutes of sequencing
 Unlike other sequencing platforms (Illumina or PacBio), overloading/saturation is good :
 Lifetime of a sequencing pore > lifetime of an empty pore
 More pores sequencing at the same time  higher throughput
Occupancy = 65% Occupancy = 80%
Occupancy =
Number of sequencing pores (Strand)
Number of available pores (Single + Strand)
should be >80%

18. 18 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
LIBRARY PREPARATION : EXAMPLE OF A 20 KB LIBRARY (BACTERIA)
DNA reparations
AMPure XP recovery 80-100%
(20 min binding/10 min elution
on a rotator)
If <80%, DNA is not pure or too large
(>150 kb)
End-Prep
Adapters ligation
2.2 µg of 20 kb sheared DNA (170 fmols)
1.5 µg of 20 kb sheared DNA (110 fmols)
1.2 µg of 20 kb sheared DNA
90 fmols loaded onto the flow cell
2.4 µg of 20 kb sheared DNA (180 fmols)
(recovery <80% : beads washed with
ABB buffer instead of 80% ethanol)
3 µg of 20 kb sheared DNA (230 fmols)
80%
92%
68%

19. 19 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
EXAMPLE OF RESULTS : MULTIPLEXING THE GENOME OF 12 BACTERIAL
STRAINS
 Throughput is stable and >10 Gb since MinKNOW 1.5
 >500 Mb of data for each sample
 Mean read lengths = 10 - 16 kb (without size selection)
Mean read lengths
Amount of data
Yield and read lengths (MinKNOW 1.5 – LSK108 kit and Albacore v1)
Yields for the last 4 runs (48h) Detail of the RUN 1 : 12.9 Gb

20. USE OF THE NANOPORE TECHNOLOGY TO OBTAIN NEW
REFERENCE GENOMES OF ARABIDOPSIS THALIANA :
INCREASING THE POWER OF GWA MAPPING
02

21. 21 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
METRICS OF THE ASSEMBLY OF A NEW REFERENCE : GENOTYPE A1-141
Raw data (2 flow cells)
(Albacore v2.0)
Reads >Q9
Corrected reads
(Canu v1.6)
Assembly
(smartdenovo)
Nb of reads 851 733 742 189 330 093 79
Max read length 223 941 67 237 56 518 13 866 816
N50 BP 14 159 14 295 15 357 8 826 176
N50 NUM 286 338 263 491 138 095 5
N90 BP 9 041 9 375 12 224 986 644
N90 NUM 628 950 575 508 285 194 18
MEAN 11 795 12 537 15 287 1 513 925
Total BP 10 046 292 165 9 305 065 534 5 046 155 201 119 600 115
A. thaliana depth 80X 74X 40X -
90% of the assembly in only 18 contigs

22. 22 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
COMPARISON OF THE NEW GENOME WITH THE REFERENCE GENOME
(COL-0, SANGER SEQUENCING)
Very good collinearity at a chromosome scale
Reference genome
Nanoporegenome

23. 23 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
ASSEMBLY METRICS FOR ANOTHER ACCESSION (SHAHDARA)
Raw data (1 flow cells)
(Albacore v2.0)
Reads >Q9
Corrected reads
(Canu v1.6)
Assembly
(smartdenovo)
Nb of reads 505 820 402 391 302 781 102
Max read length 102 801 60 457 51 666 13 404 585
N50 BP 15 074 15 345 15 112 8 793 362
N50 NUM 126 462 111 383 91 408 6
N90 BP 6 517 7 156 7 491 482 518
N90 NUM 303 804 263 228 213 870 24
MEAN 9 735 10 910 11 644 1 147 682
Total BP 4 924 331 663 4 390 258 563 3 525 820 530 117 063 664
A. Thaliana depth 39X 35X 28X -
Close metrics with 80X or 39X of data

24. 24 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
COMPARISON OF THE SHAHDARA GENOME WITH THE REFERENCE
GENOME (COLUMBIA)
High quality genome allows to detect large structural variations
≈2 Mb
Nanoporegenome
Reference genome

25. 25 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
0
5000
10000
15000
20000
25000
LONG READS POTENTIAL OF THE NANOPORE TECHNOLOGY
Unsheared plant gDNA with 1D kit (LSK108)
0
20000
40000
60000
80000
100000
120000
140000
21 783 reads > 50 kb (= 1.6 Gb)
1 962 reads >100 kb (= 0.23 Gb)
Raw data :
• Total yield = 4.2 Gb
• N50 = 39.3 kb
• Max = 291 kb

26. 26 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
LONG READS POTENTIAL OF THE NANOPORE TECHNOLOGY
Unsheared plant gDNA with 1D kit (LSK108)
Blast of a 250kb reads

27. 27 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
ACKNOWLEDGMENTS
Cécile Donnadieu, Catherine Zanchetta, Maxime Manno, Alain Roulet, Céline
Lopez-Roques, Céline Vandecasteele, Gérald Salin, Claire Kuchly, Pauline Heuillard
Christophe Klopp
William Marande, Sandrine Arribat, Hélène Bergès
Jérôme Gouzy, Sébastien Carrere, Fabrice Roux, Claudia Bartoli,
Cyril Libourel, Léa Frachon, Tatiana Vernié, Jaishree Subrahmaniam

28. 28 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf28 │ Nanopore Day Paris 2018 │ @NanoporeConf #NanoporeConf
The content contained in this presentation should not be reproduced
without permission of the speaker.
© Copyright 2017 Oxford Nanopore Technologies. The MinION,
GridION, PromethION and VolTRAX are for research use only.
THANK YOU