[A computerized numerical identification system for the diagnosis of bacteria isolated from raw animal food products].

Acceleration of bacteria identification and certain standardization of results in clinical, and also veterinary and food microbiology is enabled by use of commercial diagnostic sets and computer processing of their results. In the present paper, 80 bacterial strains isolated from animal sources in the department of food hygiene were used for assessment of two numerical identification systems (TNW and IDENTI). They were isolated from primary cultures on End's, meat-peptone and blood agar (Imuna, HI-Media) under different culture regimes (37 degrees C/24 h, laboratory temperature/3 days, 7 degrees C/10 days). A commercial diagnostic set ENTEROtest 1 and 2 (Lachema a.s., Brno, Czech Republic) was used for the purposes of diagnostics of gram-negative fermenting, oxidase-negative bacteria. Testing was performed according to the manufacturer's instructions by the culture of isolated strain. The test results were read continually after 6-hour (sugars) and 24-hour cultivation. Each strain was tested three times. Additional tests--oxidase, catalase, yellow pigment production. These commercial numerical identification systems were compared: IDENTI (supplied by Z. Svoboda, Jihlava; Lachema a.s., Brno) TNW (supplied by the Czech Collection of Microorganisms, Brno) Out of the total number of 80 isolated gram-negative bacteria 77 (96.25%) strains were, classified into the genera by both numerical systems. This result is in agreement with the percentage of identified strains by means of foreign commercial sets and numerical systems, which ranges from 93 to 99% in bacteria isolated from foods (Cox and Mercuri, 1979).(ABSTRACT TRUNCATED AT 250 WORDS)