Industry can use such models to:

  • formulate/reformulate food products
  • see if a specific recipe can confer them an intrinsic resistance to the microorganism action
  • test new technologies against traditional ones
  • design processing, storage and distribution chains
  • implement Food Safety and Management Systems

Authorities can make risk analysis and base the food safety regulations on predictions. Education can use microbiological models to make future specialists understand better how different processing technologies act on microorganisms. Companies and institutions using predictive models are minimizingthe lengthy and costly process of microbial testing.

The usefulness of predictive models determined various institutions to collect such models in specific databases.


One famous database is ComBase (Common Relational Data Base), which can be accessed via Internet at the address The work started in the beginning of 2000, when the Institute of Food Research (IFR) from Norwich, UK, developed a format for a database capable to collate models obtained in different laboratories. This format was adopted by the Food Standard Agency (FSA) from UK, which contributed with models from the Predictive Microbiology Programme, and by the United States Department for Agriculture (USDA), which contributed with models from the Pathogen Modeling Program (

Along the years, with financial support from the European Union, ComBase was enlarged with models from various European institutions. In 2006, the Centre for Excellency in Food Safety from Australia joined the ComBase consortium and the web programme ComBase Predictor was launched.

Two types of microbiological responses are registered in ComBase: growth curves and survival/inactivation curves. ComBase models had been used by SafeConsume to evaluate the risk associated with foods handled by consumers during transportation, storage and cooking.

Food Spoilage and Safety Predictor

Although a database as ComBase reduces the time spent locating models in journals and on various Internet sites, by offering them freely in one intuitive interface, several other databases for models describing faith of microorganisms in food exists too.

One of them is Food Spoilage and Safety Predictor (FSSP v 4.0), belonging to the Technical University of Denmark, one of the SafeConsume’s partners.

CB Premium

Another such database is the Computational Biology Premium (CB Premium), which supports a new era of predictive microbiology and risk-based food safety (e.g. FSMA), by providing predictive models that have been developed and validated in real commercial foods. This is the main feature that distinguishes this database from ComBase. CB Premium is owned by the University of Tasmania, produced and delivered through the Tasmania Institute of Agriculture and is available at Predictions are made under different environmental conditions for Escherichia coli, Clostridium perfringens, Listeria monocytogenes and Listeria innocua, Salmonella, Staphylococcus aureus, Vibrio parahaemolyticus, Vibrio vulnificus. Total viable counts can be also estimated.

The range of food addressed by CB Premium is quite large as it contains models for meat and meat products (ground beef, bratwurst, ground pork,fermented meat sausages), poultry meat (ground chicken, ground turkey), fish (salmon) and shellfish (oysters, prawns, snow crab, blue crab), cheese (Brie, Camembert, Cheddar, Mozzarella, Mascarpone), milk and cream, ice cream,liquid eggs, fruits (cantaloupe, honeydew, watermelon, avocado), mushrooms (button) andvegetables (broccoli, cabbage, cucumber, corn, green peas, green and red bell pepper, iceberg lettuce, spinach).

Besides basic models, CB Premium has a range of special models to be used in risk management:

  • E. coli inactivation in fermented meat
  • Perfringens predictor
  • Risk ranger
  • Pathogen growth predictor.

and an interface named Process lethality calculator.

To see how these tools for risk assessment are working, visit the CB Premium website (address given above).

CB Premium supports all lead model authors by providing free access to the database, and measures the number of times users access a specific model in order to determine its impact. Existence of such models led to a new branch of microbiology: Predictive Microbiology.