Top Banner
Breeding services Xavier Delannay
21

Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Dec 14, 2015

Download

Documents

Myra Threlkeld
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Breeding services

Xavier Delannay

Page 2: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Agenda

Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement GIS support GRSS MARS implementation at GCP

Page 3: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

14 use cases as first phase users of IBP

Beans  S. Beebe Africa: Ethiopia, Kenya, Tanzania, MalawiCassava  E. Okogbenin AfricaChickpeas  P. Gaur Africa: Ethiopia, Kenya 

Asia: IndiaCowpeas  J. Ehlers Africa: Burkina Faso, Mozambique, SenegalMaize (DT Maize)  G. Atlin Africa: Angola, Ethiopia, Kenya, Malawi ,

Mozambique, Tanzania, Uganda, Zambia , Maize (AMDROUT) B. Vivek Asia: China, India. ThailandRice (STRASA) A. Kumar Sub-Saharan Africa 

South AsiaRice (Green Super Rice) Z. Li Africa and AsiaRice (RI) MN Ndjiondjop West Africa - Nigeria, Burkina Faso, MaliSorghum  J-F Rami Africa: Ethiopia, MaliWheat (RI, India) V. Prabhu IndiaWheat (RI, China) R. Jing ChinaWheat  (ACIAR) R. Trethowan Asia: IndiaWheat (rust) S. Dreisigacker Asia: China 

Africa: Ethiopia, Kenya

Page 4: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Users / Developers Interaction

User committee set in place at Hyderabad launch meeting Difficulty to interact among scientists widely dispersed across

time zones (from California to Australia) Attempts to set up subcommittees not successful

Everyone very busy Best solution may be ad hoc teams regrouping

developers and interested users Field book tested at TL1 meeting in Madrid Optimas interaction

Critical at this meeting for users to give their inputs to developers

Page 5: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

IBP marker services In 2009, a new marker services concept was put in place that

uses established high-throughput genotyping services providers to support the projected rapid growth of genotyping needs Transition from low throughput, low capacity, public SSR

genotyping labs to high throughput, high capacity, commercial SNP genotyping services

6-10X reduction in genotyping costs Identification of breeder-friendly SNP platforms that can meet the

flexible needs of MAB applications Ability to ship leaf samples from around the world (no local DNA

extraction needed) Fast turnover to meet tight timelines for MAS and MABC projects Ability to integrate into the LIMS and informatics tools of the MBP

Page 6: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

IBP marker services Chunlin He replaced Humberto Gomez in October 2010 as lead of the

marker services and the GSS GSS consists of genotyping projects funded by the GCP to expose NARS

researchers to molecular breeding and help get them started with MB Needs managed by Theme 4, implementation by Marker Services

Marker Services provides access to genotyping services to interested researchers to help in their MB projects

The new marker services concept based on high-throughput SNP genotyping was implemented in 2010 Decision to focus on a single SNP genotyping provider (KBioscience, UK) SNP conversion to KBioscience platform well underway

GCP funds the conversion of the first set of SNPs Assays available to customers after that (average cost 12 cents/datapoint) Good set of genotypes fingerprinted as part of conversion process, good basis to

build on to understand germplasm relationships and provide foundation for wide MB use

SSR genotyping support still being provided by current labs as needed ICRISAT BecA DNA Landmarks

Page 7: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Crops Partners # SNPs Status

Maize CIMMYT 1250 Available for genotyping

CowpeaUniversity of California Riverside - Jeff Ehlers 1122 Available for genotyping

Chickpeas ICRISAT - Rajeev Varshney 2005 Available for genotyping

Pigeonpeas ICRISAT - Rajeev Varshney 1616 Available for genotyping

Rice IRRI - Michael Thomson et al. 805 Available for genotyping

Cassava IITA - Morag Ferguson, P. Rabinowicz 1740 Available for genotyping

Sorghum EMBRAPA - Jurandir MagalhaesCIRAD - Jean-Francois Rami 1578 Available by June 10, 2011

Common bean   1500 Available by end of 2011

Wheat   1500 Available by end of 2011

Available SNP Markers for Genotypinghttp://ibp.generationcp.org/confluence/display/MBP/Activity+3.1.2

Page 8: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Breeding Planning Services

Breeding schemes available on IBP wiki MAS MABC MARS

Goal to develop macros to allow calculation of costs of different breeding scenarios

Page 9: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Importance of Phenotyping Services

The GCP and the Gates Foundation are funding extensive efforts for the implementation of MB into breeding

Good sets of marker tools are now available for low cost, high quality genotyping

The generation of quality phenotypic data is a critical component of a successful implementation of molecular breeding in developing countries Need to get accurate and precise information on trait-marker linkages for

effective predictive use of markers in breeding (MAS) Precise phenotyping needed to accurately identify genomic regions of

interest for recombination in segregating progenies (MARS) Quality multilocation trials needed to assess GXE effects and help in

assessment of potential usefulness of new QTLs

Page 10: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Local Phenotyping Capacity: An Issue

In many NARS, phenotyping capacities are not sufficiently developed to face the challenges of uniform screening conditions and controlled stress environments Constraints in:

facilities and human capacity documentation and data management

Competition for good land and resources

There is a need to characterize phenotypic sites for: Climate data Soil conditions

There is a need to better integrate multi-location phenotypic data Shared genotypes and protocols, quality of data collection

Page 11: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Strategy for GCP Phenotyping Network

Shift with CI concept from a primary focus on a few centralized sites (mostly CG-managed) to the use of multiple decentralized sites (mostly managed by NARS)

Implementation strategy Complete the characterization of local sites by GIS team Identify sites in need of infrastructure improvements Establish prioritized list of needs for each year of MBP plan Use combination of MBP, TL1 and CI funds to help improve

capacity of key sites ($700K for each of first two years, lower amounts after that)

Dr. Hannibal Muhtar was hired as a consultant to help in the evaluation and the establishment of the infrastructure improvements for the African sites

Page 12: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Summary of phenotyping sites

bea

ns

cass

ava

chic

kpea

cow

pea

gro

un

dn

ut

mai

ze

rice

sorg

hu

m

wh

eat

Africa (32 sites):Benin 2Burkina Faso 2 1Ethiopia 1 1Ghana 2Kenya 1 1 2 2Malawi 1 2Mali 2 1 5Mozambique 1Niger 1 1Nigeria 2 1 1Senegal 1 1Tanzania 1 1 1Zimbabwe 1

Asia (24 sites):China 2 4India 5 1 3 2 4Indonesia 1Philippines 1Thailand 1Vietnam 1

Americas (9 sites):Brazil 1 2Colombia 4 1USA 1

Total 8 7 8 5 8 10 8 10 8

Page 13: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.
Page 14: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Summary of improvements funded in first two years of implementation

Project Location Infrastructure improvements funded

Africa:Minjibir, Kano State, Nigeria Irrigation, rainout shelter, weather stationManga, Ghana Irrigation, rainout shelter, weather station

Chickpea Egerton University, Kenya Extension of irrigation facility, weather station, fencingCowpea IIAM station, Chokwe, Mozambique New irrigation pumpGroundnut Naliendele Research Station, Tanzania Complete irrigation system

Badeggi, Nigeria Irrigation, weather station, tensiometersBanfora, Burkina Faso Irrigation, weather station, tensiometersLongorola, Mali Irrigation, weather station, tensiometersSotuba, Mali Irrigation, other site improvementsCinzana, Mali Irrigation, weather station, other site improvements

Sorghum (Al tol) Sadoré (Niamey), Niger Irrigation equipmentChepkoilel, Kenya Irrigation, fencing, greenhouseSega, Kenya Irrigation, fencing

India:Chickpea Regional Agricultural Research Station

(RARS) of ANGRAU, NandyalRainout shelter, upgrading of weather station

Indira Gandhi Agricultural University, Raipur Rainout shelterCentral Rainfed Upland Rice Research Station, Hazaribag Rainout shelter

China:Wheat Four sites Shelter for heating stress

Cassava

Rice

Sorghum (MARS)

Rice

Comparative genomics (maize and sorghum)

Page 15: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

GIS Tools (Glenn Hyman)

Improving geographic targeting Planning multi-environment trials Support GxE analysis Support phenotyping Modeling tools for phenotyping Information package for MBP trial sites

Page 16: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Genetic Resources Supply Service (GRSS)

Validation of germplasm reference sets of 19 crops continues; unanticipated delays have been experienced

Genotyping and analyses of data completed for all crops except for cassava Differences observed between the original and validation dataset, further testing ongoing.

Validated reference sets and Microsatellite Kits for sorghum and chickpea are now available from ICRISAT and CIRAD

The reference sets will be used in a pilot program to evaluate demand, protocols for maintenance, sustainability and quality assurance

Validation for reference sets of 8 priority crops (including sorghum, chickpea, maize, wheat, rice, cowpea, groundnut, and common bean) are expected to be completed by July 2011

A complete report for all expected by October 2011 A Singer-based ordering portal  for the reference sets has been developed by

Bioversity; other sets will be cataloged and accessible through the portal as they become available

Page 17: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

GCP MARS concept

MARS concept demonstrated in large seed companies (maize, soybeans) Large-scale testing needed to identify small QTL effects

MARS has great potential for many developing country programs Lower historical intensity of breeding means that large QTL effects should

still be present (low-hanging fruits) Probably fewer QTLs to recombine than for commercial programs

MARS process implemented as proof of concept for GCP crops Beans, cassava, chickpea, cowpea, rice, sorghum, wheat Optimum implementation will vary from crop to crop

Opportunity to test various options during first implementation phase

Page 18: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

MARS implementation specifics

Typical MARS program uses crosses made by breeders in their traditional breeding programs Look for good complementarity in parents Select parents of similar maturities to reduce variability in yield testing Fingerprint each parent to identify sets of polymorphic markers spread on average

every 10-20 cM Develop a population representing the maximum range of genetic variation

Generate a population of 200-300 F2- or F3-derived lines No phenotypic selection during population development, except for traits of critical

importance (MAS can be used if desired to select for those traits) Generate enough seed from each F2 or F3 plant to conduct yield trials, for

instance to F2:4 or F3:5 if two generations needed Phenotyping done with bulked final seed for each progeny In hybrid crops such as maize, use testcrosses for yield evaluation

Sample and preserve DNA from each founding F2 or F3 plant for later genotyping, or take bulk samples from later generations Genotyping can be done at any time prior to phenotyping data collection

Page 19: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Phenotypic evaluation of populations

Each population is then field tested in multiple locations appropriate for evaluation Only 1 or 2 reps needed per location, but use as many locations as possible

Goal is to identify QTLs that are significant across multiple environments (limited GxE interaction)

Very important to have quality phenotypic data (use alpha lattice or other improved design)

Use across-location average for each progeny for QTL analysis Measure as many useful traits as possible to take advantage of the

MARS process Testing for abiotic stresses will require two sets of locations

Irrigated vs. non-irrigated for drought tolerance

Page 20: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Mechanics of recurrent selection

Define sets of complementary progenies for recombination Plant out 8-10 seed of each selected progeny Genotype individual plants and select in each progeny the plants with

the best combination of chosen QTLs to recombine Cross selected plants from complementary progenies to combine their

QTLs Do in two or three stages:

A x B and C x D, then intercross F1s Select progenies with best genotypes and redo the cycle until most

QTLs have been recombined It is important to use several independent sets of plants in parallel in

this process to avoid losing too much variability at unselected loci Software will be available from the IBP to facilitate this process

Page 21: Breeding services Xavier Delannay. Agenda Use cases Users / developers interaction Marker services Breeding planning services Phenotyping site improvement.

Parent 1 X Parent 2

Po

pu

lati

on

d

eve

lop

me

nt

F1

F2

F3

F3:4

F3:5 ( if needed)

Single seed descent

300 F3 progenies

300 progenies

Multilocation phenotyping

1st Recombination cycle A B C D E F G H

F1 F1 F1 F1

F1 F1

F1

F2

F3

2nd Recombination cycle

3rd Recombination cycle

Multilocation phenotyping

F3:4

Rec

om

bin

ati

or

Po

pu

lati

on

d

eve

lop

me

nt

10 plants/family (A-H), 4 sets of 8 families/cross

Bi-parental population

QTL detection

Genotyping