To me the answer is the genes taken from a species does not match readily in most cases in the target species.

doi:10.1038/ng.3484

In this paper they make a comparison of breeding methods. This might partly answer your question.

Not exactly, if you are talking about GM breeding vs. Conventional breeding:

For example, if you want to develop a resistant cultivar. If you already have the resistant gene(s), you can just clone them and genetically transfer them into the desired cultivar (suppose that the cultivar is ok for transformation). On the other hand, you can also breed in a resistant gene from a non-elite line by conventional breeding method through crossing. But after that, you have to do many runs of backcrosses to regain the cultivar genome makeup. The backcross process may take a few years (even under marker-assistant selection). The worst part is that the 'linkage drag' may linger around forever, which might reduce the fitness of the cultivar.

Another factor you have to remember is that some of the effective foreign genes CAN NOT be acquired through same species crossing. Some of these useful genes for crop effective management are from bacteria. The examples are those herbicide- or insecticide-resistant GM crops.

The comparison is to be done based upon the objective of breeding a variety. The longer time can be taken by both the approaches depending upon the breeding objective, breeding material, availability of genomic resources, technical support etc. If you have to correct a specific defect of variety and the genes and genomic resources are known its better to go for molecular breeding for faster results. Similarly the mass selection and pedigree breeding can be easier and faster for major gene having dominant inheritance and high heritability.  

Molecular breeding is based on markers linked to phenotypic traits being transfered to elite breeding lines through a back crossing program if one wants to recover the original inbred. Foreground based selection of plants containing the marker/s under selection results in developing a value added inbred in six generations of backcrossing. Considering the fact that the trait under selection is being selected based on genotypic marker (assuming it to be single gene dominant)  field evaluation to recognise the phenotype is not required. Hence in 18 to 24 months (depending on crop & facilties to grow the plants through out the year (under controlled protected green or poly  houses)  the introgression is complete. This would take almost twice the time in conventional breeding as one would have to have the capacity to select the phenotype under conditions that help express the same. Hence round the year back crossing cant be followed. In case the trait under transfer is a recessive, once again molecular breeding allows you to select based on genotypes & no need to phenotype. Critical factor is that the molecular marker is tightly linked with the phenotype. Even under recombinant breeding to create new inbreds with the trait under selection ( pedigree breeding) the molecular approach can help select the best F2 plants carrying the gene  and having agronomic traits needed in the project and this can be followed by a SSD approach till F6 when F6 families can be opened and selected for what is aimed in the project. In this case also the time taken is almost 60% of what it will take in a conventional approach. If we are involved in transfering multiple genes or QTL linked traits, critical factor to be remembered is to raise a very large F2 progeny under conditions for which the breeding project is destined. The formula is 4^n. if four genes are under transfer 1 plant out of 256 plants will carry all 4 genes in a homozygous state. Hope this helps 

Dear Dr. Tikoo,

Thanks for the answer.

I have a few questions related to 'introgression'.

(1) When you say "in 18 to 24 months (depending on crop & facilities to grow the plants through out the year (under controlled protected green or poly houses), the introgression is complete."  Does this include 6-7 generations of backcross? What kind of crops you referred to?

(2) For a company, what kind of 'Linkage Drag' status they will 'accept' for releasing as a new cultivar, and stop more backcrossing? Some of the linkage drag can linger forever and reduce fitness for the plants.

(3) For companies, do they still do 'phenotyping' these days, or do they rely more on genotyping?

Thanks,

Yuan-Yeu

Dear Dr Yuan,

My answers to your questions

1. In a crop like tomato we go for a max 100 days of seed to seed cycle - as we just need the first fruit to ripen & harvest seeds. Since Tomato doesnt have dormancy or photoperiodic effect we sow for next back cross within a week of seed harvest. In a polyhouse we can grow a crop any time as rains or higher temperatures wont matter if it is a high quality structure.  Any similar crop (Corn in field crops is a good example)  that has a four or five month seed to seed cycle can also be used same way. As long as markers are well know & linked with phenotype quite well. In six backcrosses the trait is introgreseed with over 99 percent recovery of the recurrent parent genes. If the genome of the crop is already known, background based introgression the back cross introgression is complete even in two to three backcrosses max. 

2. Linkage drag is mainly a problem when one is using interspecific crosses. in such crosses the back cross F2 with a very large population is needed to find plants that have broken the tight linkage. again sequencing based selection can be of help. I agree that where linkage drag is existing especially if linked with deliterious genes & hence phenotypes too, one needs to spend that much more time to succeed. Commercially these are not the only projects. There are so many genes known in several crops where markers are known & need to be transfered to elite inbreds

3. Please understand that before a molecular marker is released for actual usage by breeders a detailed phenotyping is routinely done to validate the markers. In tomato over 16 such genes are today routinely used in breeding programs. 

Hope that helps. all the best

Dear Dr. Tikoo,

Thanks for the answers. Very interesting. I have a couple of follow-up questions:

1. For answer #1, I totally understand.

2. For your answer #2, you said "Commercially these are not the only projects. There are so many genes known in several crops where markers are known & need to be transferred to elite inbreds". Does that mean that the company is going to throw out this project, if the 'Linkage Drag' (with deleterious alleles) cannot be removed after even 6-9 backcrosses? What about if they are trying to introgress a VERY IMPORTANT gene? Still discontinue the project?

3. For your answer #3, I know we have to score phenotypes (traits) when looking for markers linking to them. But, what do you mean by "(In tomato) over 16 such genes are today routinely used in breeding programs"? Did you mean, if you are looking for markers to a trait (such as disease resistance), you also pay attention to the presence of those 16 extra traits (genes)?

Thanks,

Yuan-Yeu

Thanks Dr Yeu. My comments

2: I never said that the projects that have linkage drag or are more difficult are discontinued by companies. Today most of the genes for resistance in tomato have been transfered from wild species some of which have had incompatibility problesm & some linkage drags but the genes have been transfered to very successful commercial inbreds over last three cecades in the private sector. Similar is the case with many other crops and many of these even in the public sector. Unless you are specific about which project you are refering to I wont be able to comment.

3. What I meant is that if you see tomato hybrids grown by farmers globally you will see that these have traits like dis resistances:  V1, V2, I1, I2, I3, Tm2a, Mi, Ty1, Ty2, Ty3, S, BWR6, BWR 12,  Fcr, Cl1 to Cl (several genes) etc, Plant traits like  sp (determinate), ogc (colour gene), hp (colour gene), etc. All these genes are being tracked with molecular markers. Depending on what type of combination is needed in a hybrids, projects are planned & executed to deliver in very precise timelines. Rice is another big example where over 10 genes are being used in molecualr aided breeding projects by most companies having facilities to do so. We are involved in both crops in our company. 

Thanks

Suren

Thanks for the answers, Dr. Tikoo:

1. By the way, among those genes you mentioned [V1, V2, I1, I2, I3, Tm2a, Mi, Ty1, Ty2, Ty3, S, BWR6, BWR 12,  Fcr, Cl1 to Cl (several genes) etc, Plant traits like  sp (determinate), ogc (color gene), hp (color gene), etc.] in the hybrid tomatoes grown by today's farmers, is any gene transferred/introgressed by using 'Genetic transformation' method (instead of conventional breeding through crossing)? Genetic transformation technology has also been around for ~30 years.

2. Are the reasons to "why companies preferentially choose 'conventional breeding' over 'genetically transformation' to bring in important genes" due to the 'long period of time and huge of money' to de-regulate a GM product? I have heard and read that nowadays only giant companies like Monsanto can afford to do that. And the big companies are eating up the small companies. What is your opinion on this?

Thanks,

Yuan-Yeu

@  Md. Abul Kalam Azad: 

"Why molecular approach takes more time to develop a new crop variety than conventional breeding?"

Sometimes, it also takes so long to 'de-regulate' a GM variety. See the recent article from TheScientist Magazine. See the yellow highlighted part. The example used in the article, it took nearly 10 years to de-regulate the specific GM apple.