1- Scientific context and project objectives

The variability of muscular characteristics (and therefore the quality of beef) stems from the genetic polymorphisms and changes in the expression of the genome associated with different rearing conditions. Known muscular characteristics (fibres, lipids, collagen) only explain up to 30-40% of the variability of tenderness. In these conditions, new molecular predicters of meat quality need to be found.

The genome approach has been implemented to look for polymorphisms in potential genes involved in muscular biology (European programme GEMQUAL). Several hundred SNPs have been identified in 200 potential genes. These will be compared with the quality of meat in 15 breeds of cow. The QUALVIGENE project, (AGENAE programme, Analysis of the GENome of Reared Animals, GENANIMAL call for projects of the National Research Agency), has to validate this type of polymorphism, either from the GEMQUAL project, or other publications or even patents, of the breeds Blonde d’Aquitaine, Charolaise and Limousine (3,550 bull-calves). It is being conducted with the French livestock industry (UNCEIA, Institut de l’Elevage, APIS-GENE).

Likewise, transcriptome and proteome approaches have made it possible to publish lists of genes or proteins that are expressed differently depending on the quality of meat or muscle growth potential (INRA Theix; USA), the "marbled" appearance (CSIRO Australia) or tenderness (INRA Theix). The MUGENE project, conducted by the French National Institute for Agricultural Research in Theix and funded by AGENAE, aims to carry out genotyping, genome studies (transcriptome, proteome) and biochemistry so as to combine all of the results obtained. The QUALVIGENE and MUGENE projects have been officially recognised by the Inno-Viandes Competitiveness Cluster of the Massif Central and sponsored by the Meat Institute Development Association (ADIV, Clermont-Ferrand). Lastly, the genes identified by these methods are known in other animal species (rodents and humans for example) and, while the amount of information available increases very quickly (sequencing or mutagenesis programmes for example), integrating all of the data is still a major scientific challenge. The objective is to link the information available on the genes involved in determining meat quality by processing all of the data acquired on genetics, muscular biology and meat sciences to develop tools for predicting the end quality of beef.

2- Project description

This research project has two main objectives:

• Develop an information system, i.e. organise and store in suitable databases, then analyse the increasing flows of biochemical and genome data on the quality of beef
• Develop a software tool for predicting the quality of beef through modelling approaches based on research that has already been carried out in other countries (Australia, USA).

For the first objective, research is intended to be based on the GEMQUAL, QUALVIGENE and MUGENE projects. This data, on over 4,000 animals, is currently shared between the different partners through various ACCESS databases (including a base called FILICOL) that can be accessed via web sites. The aim is to gather, store, save and analyse all of the phenotype and molecular data from these programmes to identify new markers of beef quality.
For the second objective, we will look at existing muscle characterisation models, developed in the US (http://bovine.unl.edu/eng/), or meat quality prediction models based on rearing factors and techniques, developed in Australia (Meat Science, 2002, 62 :295-308), to construct a beef quality prediction model in line with the requirements of the French markets. This will be carried out in partnership with the CSIRO in Australia, with which we conducted a project called “Beef marbling”.

In practice, we will:

• Draw up a bibliography of the available mathematical methods for modelling meat quality
• Combine the available information systems (database compatibility, etc.)
• Test or develop robust database query and linking analysis tools
• Carry out mapping in silico to position the genes identified by different approaches on the beef genome (genotyping, transcriptome, proteome)
• Process all of the data obtained from the different programmes with a view to “integrating” it
• Publish the results obtained

3- Public or controlled use

The gene markers (polymorphisms, level of gene expression) and beef quality prediction tools will be made available to our partners in the French beef industry (UNCEIA, Institut de l’Elevage, APIS-GENE) and our foreign partners..

4- Expected results

The project should develop computer tools for modelling the quality of beef using phenotype data (characteristics of animals and their muscles) and molecular data (polymorphisms, transcriptome and proteome data). The results of this project in LIFEGRID will be used in the European programme PROSAFEBEEF (2007-2011) involving INRA, ADIV, APIS-GENE, CSIRO (Australia) for “quality prediction” and American universities for transcriptome. A specific Work Package focuses on predicting the quality of beef from “integrated” phenotype, biochemical or molecular data. It is coordinated by JF Hocquette (France) and involves APIS-GENE (which provides the data from QUALVIGENE), the National Food Center, and CSIRO (Australia).

LifeGrid, the regional information system