In the world of cattle breeders and math scientists, MiX99 software is a tool that everybody knows. The success product developed at Luke for breeding value estimation keeps on conquering new frontiers.
If you ask Senior Scientist Martin Lidauer what a future cow will be like, he will describe it with a huge matrix, where a milking cow is constructed of million variables, including genetic markers.
Lidauer and his team know their matrix in and out. They did not leave it to theoretical drawings on the lecture room blackboard. Breeding data and mathematical modelling were refined into a practical tool, MiX99.
“Today, Mix99 can use all genetic information that is available. It is possible to increase the accuracy of modelling of low heritability traits like health. For example, fertility of Nordic cows increases due to the genetic information we are able to use”, Lidauer says.
Leaps and challenges
Since 1990s, MiX99 has developed at the same pace with genetics and computational power. The latest big leap is modelling the rapidly increasing information from genotyped animals.
In genomic prediction, the whole genome of an animal can be explored, instead of only a set of genes. Cow genomic data is collected with SNP panel technology.
“Ten years ago, the first genomic datasets were collected from dairy cows. Each animal adds information of about 38,000 variables to our model. And soon enough, some countries would reach one million animals with recorded genomic data”, Lidauer explains.
“This amount of information is a big challenge for modelling.”
The team behind MiX99 is dedicated and persistent. Elegant and accurate are the words Lidauer uses, when he describes the work of his colleagues, Esa Mäntysaari, Ismo Stranden, Matti Taskinen, Timo Pitkänen and Kaarina Matilainen.
“Recently, we have made very nice scientific progress. The results boosted our image in the scientific community and within the industry”, Lidauer recalls, modestly.
Unknown parents, line up!
Research Scientist Kaarina Matilainen is working on the problem known as unknown parents in the cow pedigree.
“In the genetic evaluation, the pedigree data are not complete. At some point, unknown parents are arranged into groups. We noticed that these groups were not adequately taken into account in the genomic evaluation of fertility traits”, Matilainen explains.
In MiX99, the problem caused irrational results or endless progressing. Luckily, the team was not alone with the problem. Matilainen dug into the theories discovered by Professor Ignacy Misztal and found solutions from a method called QP-transformation.
“I worked on putting QP-transformation into practise for genomic evaluations. Soon, MiX99 users get more accurate results.”
Recently, Misztal praised MiX99 publicly. As it happens, Misztal is the inventor of the other leading breeding software, BLUPF90.
Open science ensures progress
Luke team is clearly famous for mastering huge amounts of genomic information. However, the expertise was not created in a vacuum. According to Lidauer, dairy breeding has always been farmer driven and therefore, very open.
“Being a scientist in this field is so much fun. There are about ten top groups in the world. We meet, discuss our problems and solutions and laugh when we realise: I can do that better than you!”
In pig breeding, for example, the field is very different. Research teams produce solutions only for the companies, not for scientific publications.
“I think the progress we see in the dairy breeding comes from this openness. Our work is financed by the industry, which expects a good product for their money. However, one company and one team can seldom solve problems as well as a whole scientific community can”, Lidauer emphasises.
The future cow
Statistics, numbers, genes… there is a great deal of data available. But what will the actual future cow be like?
In public discussion it may even seem that efficient breeding makes livestock and plants worse, not better. Hence, Martin Lidauer wants to point out that MiX99 is a tool, not the result.
“With genomics, we have more chances to focus on quality and health in a way that could not be done until now. So, I think it is possible to get better animals and plants for the whole society, not just for farmers.”
Text: Marjatta Sihvonen
Customers from Canada to New Zealand
Recently, Martin Lidauer has had little time to dive into complex matrixes. Customers, the current and future users of MiX99, need him.
“Last year, we made several new contracts. Luke’s account managers give the frames for the negotiations, but customers also need to discuss the research they need and all details of MiX99. So, we are naturally involved in the process”, Lidauer says.
Typical users of MiX99 are the national cattle breeding associations around the globe. Among the newcomers are Canadian Dairy Network, which is responsible for all genetic evaluations of dairy cattle in Canada, and two organisations, New Zealand Animal Evaluation Limited and Beef+Lamb New Zealand Genetics, which cover the dairy cattle, beef and lamb in the country.
Long term partnerships have been established around Europe, Ireland, Spain, Great Britain, and in the Nordic countries.
Also, it is not to be forgotten that MiX99 can be used to study other organisms than cows, too. The closest partner, Boreal Plant Breeding is just next door from Luke.