[To contribute to this conference, send your message to biotech-room2@mailserv.fao.org. For further information on the Electronic Forum on Biotechnology in Food and Agriculture see Forum website.
Note, participants are assumed to be speaking on their own behalf, unless they state otherwise.]

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 09:00
To: 'biotech-room2@mailserv.fao.org'
Subject: 8: Indicators of utility of MAS in plant breeding

This is from C.R. Bhatia, India. I am a plant geneticist and breeder by profession. I have used isozyme, protein and molecular markers in genetic studies. I was a co-author in an early review on marker-assisted selection in crop plants (Molecular Breeding 1997 Vol. 3: 87-103). I have retired from field and bench research but continue international and national consultancy in agricultural biotechnology.

Whenever a new tool or technique comes up a lot of hype is generated by the developer that often reflects the vendor bias. Among the users, some look at the new tool in a highly critical mode, others are over appreciative, and yet others wish to use it as a contemporary or "fashionable" tool. When resources are a constraint, as they are in most plant breeding programs, it is important to have "reality" check based on evidences or indicators. Of course, selection of the indicators could also be biased. The utility of a tool/technique in plant breeding can be judged on the following indicators at different stages of the breeding program:

Many participants would very much like to see evidence-based answers to the points raised above, also for the selections based on quantitative trait loci (QTLs). Colleagues from the developed countries who have used MAS for a longer period may share their experiences.

Another question often asked by the breeders is: Can the MAS-related analytical work (when the molecular markers are well established) be outsourced to a private laboratory? Related to the above is: Would private investment in a service laboratory providing marker analysis commercially generate adequate returns? Can the service be provided at a cost which breeders can afford to pay?

C. R. Bhatia
17 Rohini, Sector 9-A,
Vashi,
New Bombay - 400 703,
India
neil (at) bom7.vsnl.net.in

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 09:06
To: 'biotech-room2@mailserv.fao.org'
Subject: 9: Cost of MAS - the biggest barrier

My name is Bert Collard. My background is in marker development for disease resistance traits in chickpea and wheat in Australia.

I think that the cost of MAS is the most important limitation for utilisation in developing countries. I completely agree with the previous contribution by Daniel Gianola (Message 6, November 18) that rigorous cost-benefit analysis is of the utmost importance before even considering to use markers for MAS (in any country), because this varies considerably for different species, traits, labour costs etc. Such planning should also be done from an interdisciplinary perspective: by plant breeders, molecular biologists, physiologists, pathologists and economists etc, which in my experience, is not done often enough (this was also mentioned by Hayde Galvez in Message 5 (November 18)). However, given the relatively cheap labour costs and high costs for reagents, consumable products and equipment for MAS in developing countries, I suspect that there would not be many cases where MAS would be cost effective. (I would be very grateful if anyone can suggest additional references on cost-benefit analysis).

I would also like to emphasise that the costs associated with developing markers can also be extremely high - far greater than the costs associated with MAS, yet this is often neglected in reports or discussions about MAS (even if equipment has already been obtained; if a lab requires the purchase of equipment, then the initial costs could truly be 'massive'). Furthermore, marker development can take a long period of time - probably longer than previously thought. This can offset major time-savings gained by using MAS compared to conventional breeding.

What follows on from the costs associated with marker development or MAS is where funding will come from, given that funding for agriculture in the public sector in developing countries is limited, and already stretched. Moreover, could these funds - obtained from governments or external funding bodies - maintain a marker development/MAS facility in the long term?

Unfortunately, in my opinion, I doubt that adequate funding could be provided in the majority of cases. Therefore, I question whether MAS is generally feasible in the majority of labs/plant breeding institutes in developing countries, with the exception of Consultative Group on International Agricultural Research (CGIAR) centres and some National Research Centres.

Perhaps another exception could be the use of publicly available markers, if these markers were associated with high priority traits for relevant crops in developing countries. However, this would only be possible in major crop species, and not possible with the majority of minor crop and vegetable species that are vital for many subsistence farmers (due to a complete lack of marker research in some species).

In response to a previous contribution, I have encountered the following papers that have conducted cost-benefit analysis of MAS in crop species:

Dr. Bertrand Collard
Postdoctoral Research Fellow
Department of Biological and Physical Sciences
University of Southern Queensland, Australia
Toowoomba 4350
Tel: +613 4631 2699
Fax: +613 4631 1530
collardb (at) usq.edu.au

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 10:43
To: 'biotech-room2@mailserv.fao.org'
Subject: 10: MAS for tree crop improvement and transgenics

I agree with E.M. Muralidharan (Message 7, November 18) that MAS has best contribution for perennial crops like fruit trees. Generations of breeders may be needed before an improved forest tree cultivar could be released. In our coconut tree planting program, DNA markers have been successfully used to assist the selection of parental trees (seed nuts) for the nationwide coconut replanting.

I would like to add to message 7 that MAS is very well very useful also in the production/release of a trangenic crop variety. Starting from the verification of the transgene to molecular characterization of transgenic plants and establishment/introgression into other breeding materials, MAS is the only way possible to track the 'movement' of the transgene in the gene pool.

Hayde F. Galvez
Molecular Genetics Laboratory
Institute of Plant Breeding
University of the Philippines Los Baņos
The Philippines
hgalvez_4031 (at) yahoo.com

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 11:34
To: 'biotech-room2@mailserv.fao.org'
Subject: 11: Areas that need to be debated - Plant breeding

This is from P.M. Priyadarshan, a plant breeder at the Rubber Research Institute of India.

It is indeed heartening to note that FAO has taken the initiative for an e-mail conference on the current most relevant subject in plant breeding - MAS. MAS is indeed a technique that can assist plant improvement, especially when the population is large enough and when phenotypic selection becomes difficult. Also, in crops like rubber, where the gestation period is 7-8 years, MAS can help (at least theoretically) to have a quick and routine selection either for monogenic or polygenic traits. But there are areas that need to be debated falling under this purview:

1) As pointed out by Dr. Denis Murphy (Message 1, November 17), MAS cannot be considered as a direct alternative to upstream technologies.

2) It is widely believed that quantitative trait loci (QTLs) governing yield can be 'tied' through markers. Yield is a multi-channel end point and several genes, in conjunction with the environment, will lead to high or low yield. When we talk of yield, what about multitude of environments prevailing across a country?

3) We must understand/analyze pathogen strain production. Why have so many biotypes and pathotypes originated over the years? When agriculture began centuries back, there were not much causal organisms. Or, such organisms could not get hosts to grow with. Now, with large areas under monoculture, infectious organisms have their ride to live in. Remedy to this is only and only crop diversification in all agriculturally suitable environments.

4) Cost effectiveness of MAS compared to conventional breeding is yet another aspect that needs attention. If MAS is not getting cheaper, developing countries cannot use them judiciously. A centralized facility for routinely doing MAS, either under a global entity like CGIAR or at individual national level, shall be a remedy.

5) During the 1980s, the journals of plant breeding were dominated by papers on tissue culture. The 1990s witnessed an overwhelming amount of papers on RAPDs, RFLPs and microsatellites and towards 2000 and beyond the journals were flooded with papers on MAS and molecular linkage maps. Many more molecular techniques are to come. But sadly, these technologies have not come out in a big way to boost conventional plant breeding efforts, as against the popular expectations, and have largely remained in papers. Micropropagation in woody species is yet to boom, transgenic cotton in India faced a miserable equation, and the stories and reports are many that need to be analyzed to mitigate doubt in the public about the utilization of biotechnology.

6) In India, the financial outlay (annual) for biotech research is nearly Rs. 150 crores [roughly 33 million US dollars - 1 crore is 10 million rupees, roughly 45 rupees per US dollar...Moderator]. We have been actively pursuing research for over 17 years. Theoretically speaking, biotech as such, is expected to accelerate the pace of breeding new varieties. Yet, I am doubtful how many varieties have been developed and released to farmers.

7) Environmental sensitivity in expression of genes is a crucial factor governing the expression of genes identified. The expression of genes for resistance is highly environment sensitive.

8) Man cannot duplicate the job of nature, for the nature knows all and everything. Instead of competing with nature in its processes/selection, man should try more to understand and learn the intricacies of nature, and put them into practice. Learning more about nature will sustain us to live in sustenance. Let us ask the nature to undertake the selection process, for the nature only knows how, when and what to select.

P.M. Priyadarshan
Plant Breeder,
Rubber Research Institute of India,
Regional Station, AGARTALA - 799 006,
India
Tel: Off : 91-381-2355287/2355143 - Extn:205
Tel: Resi : 91-381-2354325
Fax: 91-381-2353149
pmpriyadarshan (at) hotmail.com
alternate e-mails: pmpriyadarshan (at) rediffmail.com pmpriyadarshan (at) yahoo.co.in
personal web page: www.freewebs.com/pmpriyadarshan/

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 11:43
To: 'biotech-room2@mailserv.fao.org'
Subject: 12: Re: Indicators of utility of MAS in plant breeding

This is from P.B. Kirti. I am a Professor in the Department of Plant Sciences of the University of Hyderabad, interested in the genetic transformation of legumes for disease and insect resistance.

Of the so-far-received opinions on MAS, the opinion expressed in Message 8 (November 19) by Dr. C.R. Bhatia, Former Secretary, Department of Biotechnology, Government of India, appears to be highly pragmatic. MAS is a highly fashioned and expensive technique, that is supposed to improve the selection efficiency in breeding programs and shorten the time taken for attaining the expected goals. There is no comprehensive treatise available so far on MAS in terms of successful utilization of the technique in the development of new crop varieties or breeding material, that is undergoing active consideration. Such information is most welcome in arriving at a conclusion whether the technique under consideration is good to be pursued, particularly in plant breeding programs in developing countries.

P.B. Kirti,
Department of Plant Sciences
University of Hyderabad, Hyderabad 500046
India
E Mail: pbksl (at) uohyd.ernet.in

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 13:34
To: 'biotech-room2@mailserv.fao.org'
Subject: 13: MAS for livestock in developing countries

From Dirk-Jan de Koning, Department of Genetics and Biometry at the Roslin Institute (Scotland, UK). My area of work is design and analysis of QTL experiments in livestock.

Improvement of livestock productivity in the developing world through breeding should be focused on animals that are tolerant against a variety of diseases and can cope with the environmental stress (heat, drought).

As a consideration for the cost-benefit analysis: for animal trypanosomosis, the annual cost of curative and preventative treatments for African livestock is estimated at 35 million US Dollars (USD) with a total annual cost exceeding one billion USD (Kristjanson et al. 1999. Agricultural Systems 59:79-89). At this moment, the lack of routine recording of production and health traits, alongside limited availability of molecular research facilities to individual countries (as pointed out by Olusola Sokefun in message 3, November 18), prohibits the implementation of any structural breeding program exploiting MAS. [Animal trypanosomosis, caused by trypanosomes, protozoan parasites, transmitted by the tsetse fly, is an important constraint, if not the most important constraint, to livestock and mixed crop-livestock farming in tropical Africa. The paper referred to is by P.M. Kristjanson et al, entitled "Measuring the costs of African animal trypanosomosis, the potential benefits of control and returns to research". Among other things, their results indicate that the potential benefits of improved trypanosomosis control, in terms of meat and milk productivity alone, are $700 million per year in Africa and that the disease costs livestock producers and consumers an estimated $1340 million annually, without including indirect livestock benefits such as manure and traction...Moderator].

However, MAS can be an important tool in the development of more robust breeds that can cope with the biotic and abiotic stresses of the (sub-)tropical climates in many developing countries. In agreement with previous messages, I feel that this should be done as large international collaborative efforts and not by individual countries. In such a scenario, existing breeds are assessed for their tolerance and production and subsequently a variety of experimental crosses can be made. QTL analysis followed by MAS would be used to derive a robust synthetic hybrid from the experimental crosses. Given the international nature of such an approach, any breeds would need to suit the particular challenges of each of the participating countries. The logistic and social issues related to introduction of a new hybrid breed for smallholder farms pose additional challenges for such an effort. International collaboration is a must for both the developed and developing world because studies by individual groups can suffer from limited statistical power, thus prohibiting a clear picture of the overall genetic architecture of the trait of interest.

The laboratory cost of MAS may be 'massive', but so is running a structured breeding program: At the workshop on MAS held in Turin on 17-18 October 2003 [proceedings available at http://www.fao.org/biotech/Torino.htm ...Moderator], we were told that the Italian government contributes annually about 93 million Euro towards the breeding value evaluations in Italian livestock. In plants, the development of a new cereal variety was estimated between 1 and 5 million Euro. Many developed and developing countries have large and efficient genotyping facilities available that could be targeted towards livestock improvement in an internationally coordinated program.

Dirk-Jan de Koning
Roslin Institute
Genetics and Biometry
Roslin, Midlothian, EH25 9PS
UK
phone:+44 (0)131 5274460
fax: +44 (0)131 4400434
e-mail: DJ.DeKoning (at) bbsrc.ac.uk

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 13:56
To: 'biotech-room2@mailserv.fao.org'
Subject: 14: Crossing-over // false-positive markers

This is from Adilson Mota, a research molecular geneticist working at the National Dairy Cattle Research Center of Embrapa (Brazilian Agricultural Research Corporation), located in Juiz de Fora, state of Minas Gerais, Brazil. Recent research activities focus on functional genomics related to bovine mammary gland physiology and pathology.

I have read all messages so far with great interest. But I have other questions that I would like to address. Besides the economics/labor involved in MAS (regarding construction of reference population, genotyping, data collection, and analysis), the selected marker may even, at the end, be far enough from the QTL (or gene). In this case, crossing-over(s) in that region may cause chromosome recombination and selection based on the marker genotype (MAS) may not be effective. There is also the possibility of generating false-positive markers, depending on experimental design and analysis, for which MAS should be preceded of evaluating the marker in an independent population.

I would appreciate hearing the experiences/opinion of participants regarding these (and other) technical issues of MAS and get some recent references on these subjects. I also hope that, at some point, the discussions will consider MAS related to livestock genetics.

Adilson Mota, PhD
Embrapa Gado de Leite - Dairy Cattle Research Center
R. Eugenio do Nascimento, 610
Dom Bosco - Juiz de Fora - MG
36038-330 - BRAZIL
Phone (+55) (32) 3249-4825
Fax (+55) (32) 3249-4701
E.mail amota (at) cnpgl.embrapa.br
HTTP://WWW.CNPGL.EMBRAPA.BR
HTTP://WWW.EMBRAPA.BR

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 15:57
To: 'biotech-room2@mailserv.fao.org'
Subject: 15: Re: Crossing-over // false-positive markers

Responding to Adilson Mota (Message 14, November 19), both problems you raised are important, and there are no easy answers. With respect to "false-positives", which is a hugh problem for whole genome scans, the best solution to date is the "false-discovery-rate" (FDR). The FDR does not give a definitive answer as to which effects are real, but you can at least get an indication if there is something really there. With respect to the question of QTL location with respect to the markers, confidence intervals will be large, unless both the QTL effect and sample size are large. Genotyping additional markers in the region of interest doesn't help much. Once you are down to a marker spacing of 10 cM, additional markers will not reduce the confidence interval.

Joel Ira Weller
Institute of Animal Sciences
A. R. O., The Volcani Center
P. O. Box 6
Bet Dagan 50250
ISRAEL
E-mail: weller (at) agri.huji.ac.il
Phone : 972-8-9484430
Fax : 972-8-9470587
http://www.agri.huji.ac.il/~weller
http://www.agri.gov.il/People/JoelWeller.html

-----Original Message-----
From: Biotech-Mod2
Sent: 19 November 2003 17:03
To: 'biotech-room2@mailserv.fao.org'
Subject: 16: Re: Crossing-over // false-positive markers

From Dirk-Jan de Koning, UK.

A very specific reply to Adilson Mota (Message 14, November 19) about false positives and MAS.

There is a very good paper on this issue by Hayes and Goddard (Livestock Production Science Volume 81, Issues 2-3, June 2003, Pages 197-211), entitled "Evaluation of marker assisted selection in pig enterprises". In a nutshell (at the risk of misrepresenting some of the article): When you use a very relaxed threshold for quantitative trait locus (QTL) detection, you detect (and use) many QTL that are false positives BUT you also detect most of the REAL QTL, hence increasing your genetic gain! The profit (yes, it includes a cost benefit analysis) becomes then a function of the QTL effects and the genotyping costs. I realise that Livestock Production Science is not widely available but you can always request a copy from one of the authors (ben.hayes (at) akvaforsk.nlh.no). To me, it was a refreshing read compared to all the literature claiming we need highly stringent genome-wide significant thresholds. I guess the latter depends on how damaging the detection of a false positive is. I agree with Joel Weller (Message 15, November 19) in his endorsement of the false-discovery-rate (FDR) for getting an idea on the expected number of false positives among your QTL.

Dirk-Jan de Koning
Roslin Institute
Genetics and Biometry
Roslin, Midlothian, EH25 9PS
UK
phone:+44 (0)131 5274460
fax: +44 (0)131 4400434
e-mail: DJ.DeKoning (at) bbsrc.ac.uk