BEIJING — Xiaoxiang Hu, Meiying Fang, and Jianfang Liu are all professors at China Agricultural University, where they focus on using genomic tools to improve different livestock: chicken, pigs, and dairy cattle, respectively.
Despite the differences in their research subjects, all three researchers in recent years have adopted SNP genotyping arrays to further their studies, and none is hesitant to stress the benefits of using the technology. In Hu's words, the advent of SNP arrays in agricultural biotechnology has been nothing short of a "revolution," cutting the time needed for marker discovery from years to months. Liu, for his part, says that arrays are now the "most state-of-the-art technology in animal breeding as well as in plant breeding." And all three researchers have been working with partners in China's agricultural sector to translate their findings into selection programs.
BioArray News spoke with the scientists during a visit last week to CAU's State Key Lab for Agribiotechnology. Below is an edited transcript of that discussion.
How has the availability of SNP arrays impacted your research?
Xiaoxiang Hu: Before I adopted SNP arrays, I used microsatellite markers to perform genotyping and to perform linkage mapping for interesting genes. If you used microsatellites, it would take maybe one or two years to have all the markers genotyped, but with a SNP array, in just one or two weeks you can get all of the genotyping results. And the mapping results are more precise than ever. You can use these arrays to do association studies and to do linkage mapping. Before we used the SNP arrays, it would take between five and seven years to do fine mapping of genes in, say, influenza. But with SNP arrays, it now only takes about half a year to get a final result, and this result is more precise. I think it's at least ten times faster than it was before.
Jianfang Liu: Using SNP arrays we can do a lot of things that could not be done using microsatellites, such as genomic selection. We can use high-density SNP chips to estimate genomic breeding value, because high-density SNP markers can catch the linkage disequilibrium between markers and the genes. Such LD can be captured by high-density SNP arrays, but it cannot be done by microsatellites. So, for genomic selection, this is the most state-of-the-art technology in animal breeding as well as in plant breeding.
Hu: Before, with microsatellite markers, genome selection was not possible. But using SNP arrays it is possible to perform genomic selection in animal breeding. For us, it's been like a revolution.
Why is there a need to do this kind of research?
Liu: I think there are two aspects. One is that we need to increase productivity, such as milk production or meat production or egg production. We need to find genes affecting these traits, and based on mutations in these genes we can select the favorite allele to include in our animal production. Another goal is to find genes affecting disease resistance to try to tackle the problem of disease.
Meiying Fang: Also, we can decrease the generational intervals. Usually with a cow you have to wait four to six years before you can do genomic selection based on phenotype, but with arrays we can decide earlier on which animals to breed, reducing the cost and increasing production.
Could next-generation sequencing also be used in your research?
Hu: We already have a lot of NGS applications in our research. We have a lot of projects focused on investigating polymorphisms and mutations in different genes in different chicken breeds. I think using NGS in genomic selection is the way forward. At the same time, I think the biggest bottleneck for NGS is the time for bioinformatics analysis and for data mining. With NGS, you can generate genomic data in a very short period of time, but with the amount of data it will take time to perform bioinformatics analysis. In our university we don't have a lot of people focused on bioinformatics. We need to collaborate with companies like BGI to perform data mining and it takes a long time. But this is an example of why SNP array data can be more useful that NGS data. When you need to have genotyping done in a very short amount of time because the period for selection in, say, pig is only two or three months, you need to get all of the genotype data and make a decision on which animal to breed for the next generation. But with NGS, it is not possible to do that. By the time we get an answer, several generations will have passed. So I think in the future, high-density arrays should always be good for genomic selection, but using NGS you can sequence more population founders and get custom design arrays mainly focused on the population you are breeding. So [future] chips will become more and more informative than the commercial chips that are [currently] available. For breeding companies, if the money is available and the arrays become cheaper, and more custom designs are available, combined with the NGS technique, the SNP array is a good choice.
Dr. Liu, there are a number of bovine arrays on the market today. Which one do you prefer?
Liu: I think the Illumina BovineSNP50BeadChip is good. The 100K is not so popular in China. So far we have just genotyped about 100 individuals with the 700K just to impute those with 50K to get the higher-density SNP data without doing extra genotyping because of the price.
One issue that is raised is that genomic selection could affect genetic diversity down the road. How do you think that could be avoided?
Fang: We have a strategy for this kind of problem. We maintain one population as a conserved population, so that way we keep all the traits, even if at the moment they are not so good, and then we have the commercial populations. And in this way we can avoid this issue.
Liu: We can also breed different lines. For each line we can focus on one trait to capture a certain performance and in another line another trait, and later we can cross the lines if needed.
Hu: I actually think it is difficult to avoid this issue. Commercial farms tend to care more about profit than genetics. But for animal genetic resources, the government should pay more attention. At the national level the government should pay to conserve populations.
How would you compare the adoption of SNP array and NGS technology for agbio research compared to, say, the US?
Liu: Well, at the technical level, all of the companies with SNP array technology — Illumina, Affymetrix — are in the US. In this regard, we in China are following the US.
Hu: At the same time, in China we have a lot of genetic resources. These resources are unique, so I think from the technical viewpoint we are in the same position as in other countries, and in the research area we are in the same neighborhood as well.