snp discovery from amphidiploid species and ... · • m2p = 1*g • snp score = 2 3. no conflict...

1

SNP discovery from amphidiploid species and transferability across

the Brassicaceae

Jacqueline BatleyUniversity of Queensland, Australia

[email protected]

2

Outline

• Objectives• Brassicas• Genome Sequencing• SNP discovery• SNP validation• Cross species transferability• Application• Future work

Objectives

• Development of bioinformatics tool for SNP discovery and annotation

• Establish cost effective discovery and validation of SNPs within the amphidiploid B. napus

• Assess association of SNPs with genes for agronomic traits

• Assess the extent of LD within B. napus• Assess genetic diversity of important agronomic genes

within cultivated Brassica spp. and wild relatives• Establish a strategy for SNP discovery from other large

and complex genomes.

3

Methodology

• Paired end sequence from parents of mapping populations

• SNP discovery• Genotyping using golden gate and infiunium

assays• SNPs genetically and physically mapped• Cross species amplification to other

Brassicaceae members

4

Brassicas

Diversity genomics

• Characterising genomic and phenotypic diversity in cultivated and wild plant species and their pathogens� Brassicaceae, Leptosphaeria maculans

• Investigating genetic variation in crops and wild relatives

• Investigating the evolution of plant pathogen interactions

• Identifying novel genes and genetic markers for traits of interest, such as disease resistance

6

7

Genetic diversity

• Germplasm collections are valuable gene pools • Assessing genetic and genomic diversity within these

collections:� assign lines and populations to diverse groups� study the evolutionary history of wild relatives� verify pedigrees and fill in the gaps in incomplete pedigree or

selection history� monitor changes in allele frequencies in cultivars or populations� help narrow the search for new alleles at loci of interest.

8

Domestication bottlenecks

• B. napus canola, B.juncea mustard and B. carinata are allopolyploids.

• Rare natural polyploids only incorporate a limited genetic diversity from progenitor diploids.

• Wide genetic diversity in B. rapa, B. nigra, B. oleracea progenitors and wild relatives, options to enhance canola and mustard.

• A range of strategies is available to realise the genetic potential of the Brassicaceae.

Sequence data

• Illumina GAIIx and Hi-Seq data for:• 8 B. napus cultivars• 2 B. rapa cultivars• B. oleracea• 3 Brassicaceae

• Funding for 100+ Brassicas

9

10

Brassica genome sequencing

• B. rapa ssp. Pekinensis var. Chiifu• 10 chromosomes, ~550 Mbp• Multinational Brassica genome sequencing

committee originally agreed BAC by BAC sequencing approach

• >100,000 BAC end sequences• >600 BACs sequenced• Genome sequenced using Illumina GAIIx

B. rapa SNP discovery and genotyping

• Illumina paired end sequence from parents of mapping populations

• SNP discovery• Genotyping using golden gate• Physical mapping• Cross species amplification to other

Brassicaceae members

11

SNP validation

Genotyping

• Illumina Golden gate system• 384 SNPs

• 2 B. rapa mapping populations• Parents of B. napus mapping populations• Selection of wild Brassicaceae

13

SNP Validation

SNP Pool 3Lenient Criteria

SNP Pool 2Less strict

Criteria

SNP Pool 1Strictest Criteria

~ 320SNPs

~ 50SNPs ~ 15

SNPs

GoldenGate Oligo Pool

SNP Validation

SNP Pool 3Lenient Criteria

SNP Pool 2Less strict

Criteria

SNP Pool 1Strictest filtering Criteria

94% conversion

80% conversion

30% conversion

SNP Genotyping

16

SNP Genotyping

17

18

Genetic diversity

• Assess relationships within the Brassicaceae• Correlate this with morphological and inter-

specific hybridisation data

Brassicaceae diversity

B. napus SNP discovery

• Custom algorithm developed for SNP discovery from Illumina data for amphidiploid species

• Distinguish between inter and intra genomic SNPs

22

The SGSautoSNP algorithm

• We do not consider the reference in SNP discovery• the reference is only used to bring the reads together• SNPs are called from these reads

=> different to most other SNP callers

1. coverage must be at least 42. SNP score must be at least 2

– Example:• SP1 = 6*A• AP1 = 1*G• M2P = 1*G• SNP score = 2

3. no conflict within a variety– i.e. all bases in each cultivar must be the same– if e.g. Junior 3 * A and 1 * T => conflict

23

Output visualisation


• Custom algorithm developed for SNP discovery from Illumina data for amphidiploid species

• Distinguish between inter and intra genomic SNPsXA_0011r 1252 1252 3 S=G=2;M1=G=3;Sr=X=0;A=G=3;J=T=3;M2=G=1;Bn=X=0;E=X=0; T;G;XA_0011r 1379 1379 5 S=T=2;M1=T=3;Sr=X=0;A=T=1;J=C=3;M2=X=0;Bn=X=0;E=C=2; C;T;XA_0011r 2036 2036 4 S=G=1;M1=G=2;Sr=X=0;A=G=1;J=T=8;M2=T=3;Bn=X=0;E=T=6; T;G;XA_0011r 4921 4921 2 S=X=0;M1=X=0;Sr=X=0;A=T=8;J=X=0;M2=X=0;Bn=X=0;E=C=2; C;T;XA_0011r 5070 5070 4 S=X=0;M1=G=2;Sr=X=0;A=G=2;J=A=6;M2=X=0;Bn=X=0;E=X=0; A;G;XA_0011r 5273 5273 3 S=C=4;M1=C=5;Sr=X=0;A=C=6;J=G=2;M2=X=0;Bn=X=0;E=G=1; C;G;XA_0011r 5442 5442 8 S=T=1;M1=X=0;Sr=X=0;A=T=7;J=C=5;M2=X=0;Bn=C=1;E=C=3; C;T;XA_0011r 5512 5512 7 S=G=3;M1=G=3;Sr=X=0;A=G=5;J=A=4;M2=X=0;Bn=A=2;E=A=1; A;G;XA_0011r 5976 5976 11 S=T=8;M1=T=1;Sr=X=0;A=T=2;J=C=6;M2=X=0;Bn=C=2;E=C=3; C;T;XA_0011r 5992 5992 10 S=A=9;M1=A=1;Sr=X=0;A=A=3;J=G=5;M2=X=0;Bn=G=2;E=G=3; A;G;


Base Change Type Number

A>G transition105045

C>T transition105513

A>C transversion42480

A>T transversion49287

C>G transversion29828

G>T transversion42217



A>G transition105045

C>T transition105513

A>C transversion42480

A>T transversion49287

C>G transversion29828

G>T transversion42217


A>G transition 24207

C>T transition 24375

A>C transversion 10158

A>T transversion 12254

C>G transversion 6621

G>T transversion 9918

B. napus SNP density

0

5

10

15

20

25

30

0 100000 200000 300000 400000 500000 600000 700000

Series1

B. napus SNP validation

• 24/25 SNPs correctly predicted through validation by PCR and sequencing

• 20/22 SNPs correctly predicted through Golden gate

• Range of sequence coverage and confidence scores

28

29

Gene discovery

• Finding the genes for the traits

• Integration of genetic data with genomic data• Mapping of QTL regions to genomic data

...Annotation

30

Gene discovery - application

Genetic map

Physical map

Physical scaffolds

Na12 E11 BRAS023BRMS040

BRMS005

CB10278 BRMS075

KBRH143H15

BRMS036Na12 A02

RA2 A05

1Mbp

CB10439

10 cMOI09 A06

Scaffold and Marker Assembly

Chromosome MarkerScaffoldA7

CMap3D

32Duran et al. (2010) Bioinformatics 26: 273-274

Identification of Candidate Blackleg Resistance Genes

TNL (Gene number) Scaffold1 32 33 34 35 36 37 128 129 1210 1211 312 313 314 315 1216 1217 1918 1919 1920 1921 19

TNL6 Sequence and Protein Alignment

B. rapaB. napus

1B. napus

2

B. rapaB. napus

1B. napus

2

B. rapaB. napus

1B. napus

2

Gene Mutation Species Predicted Number of Reads Sequence Verified

TNL 1 18,240Reference: B. rapa G N/A

B. napus 1 G 3 G

B. napus 2 C 1 C

TNL5 5,208,963Reference: B. rapa C N/A

B. napus 1 C 1 C

B. napus 2 T 4 T

TNL 5 5,209,056Reference: B. rapa A N/A

B. napus 1 G 1 G

B. napus 2 A 5 A

TNL5 5,209,772Reference: B. rapa A N/A

B. napus 1 A 1 A

B. napus 2 T 6 T

TNL5 5,207,023Reference: B. rapa G N/A

B. napus 1 T 4 T

B. napus 2 G 1 G

TNL 6 5,891,882Reference: B. rapa T N/A

B. napus 1 T 4 T

B. napus 2 C 3 C

Change in charge was the most common change due to protein differences

Gene discovery

Gene/EST

Primer

genomic sequence

PCR

Known Unknown(Arabidopsis) (Brassica)

http://flora.acpfg.com.au/tagdb/

39

http://flora.acpfg.com.au/tagdb

Marshall, D.J., et al. (2010) Plant Methods. 6:19

40

TAGdb output

Sym genes

Brassicas can not form symbiotic associations with rhizobia or mycorrhizae - BUT - contain homologues for many genes involved in these processes.

•What is the diversity of and selection pressure on these genes across the Brassicaceae?

•What are these proteins doing? – general pathogen/microbial perception and response?

Ferguson et al., 2010

Tagdb results

e.g. LjPOLLUX

e.g. NFR1NFR5

9 Arabidopsis homologues

e.g. LjNUP85, LjNUP133

Sequencing SYM genes in Brassicas: NSP1 (Nodulation Signalling Pathway1)


BrNSP1 and BoNSP1 vs MtNSP1 = 57% CDS similarityAtNSP1 vs MtNSP1 = 58% CDS similarityBrNSP1 vs AtNSP1 = 83.8% CDS similarityBoNSP1 vs AtNSP1 = 83.7% CDS similarityBrNSP1 vs BoNSP1 = 98% CDS similarity



High conservation in the GRAS domain.

Residues important for NSP1 function in Lotus japonicus are conserved in the Brassicaceae.



BrNSP2 vs BoNSP2 = 98% CDS similarityBrNSP2 vs AtNSP2 = 78.2% CDS similarityBoNSP2 vs AtNSP2 = 78.5% CDS similarityBrNSP2 and BoNSP2 vs MtNSP2 = 55% CDS similarity



Alanine residue important for NSP1-NSP2 interaction in Lotus japonicus is not conserved in the Brassicaceae, but conserved in rice. Rice NSP1 and NSP2 are functional in nodulation in transgenic Lotus japonicus.

Sequencing SYM genes in Brassicas: POLLUX


• One copy on both the A and the C genomes: BrPOLLUX (A), BoPOLLUX (C) – 98% similar.

•BrPOLLUX CDS is 69.4% similar to Lj POLLUX CDS, Bo POLLUX = 69%, AtPOLLUX = 61%.

•85.6% similarity between BrPOLLUX and AtPOLLUX, 85.4% between Bo and At.

•Currently sequencing POLLUXin other Brassicaceae members.

•Least similarity in N-terminal transit peptide.

Consistent with cation channel function:

POLLUX

Geneious Pro Transmembrane Prediction (Biomatters).

49

Future work

• SNP identification and genotyping of cultivated and wild Brassicaceae

• Large scale SNP discovery and genotyping for fine mapping and LD studies

• Identify which Brassicaceae to sequence • Use next generation sequencing data, molecular

markers and morphological variation to study diversity across Brassica species and wild relatives

50

Summary

• Next generation sequencing data is suitable for gene, promoter and SNP discovery in non-sequenced and orphan species

• SNPs can be applied for gene discovery and evolution in crop species and wild relatives

• High throughput genotyping can be used for fine mapping and LD studies

Acknowledgements

Emma CampbellChristina DelayMegan McKenzieReece TolleneareJoanne McLandersManuel ZanderAlice Hayward

Paul Berkman Chris DuranKaitao LaiMichal LorencSahana ManoliAdam SkarshewskiLars SmitsJiri StillerDavid Edwards

Bob ReddenHarsh RamanXiaowu Wang

snp discovery from amphidiploid species and ... · • m2p = 1*g • snp score = 2 3. no conflict...

Documents