Background:

Food composition tables become obsolete due to new production practices, plant genetic research, changes in analytical methods, and the characterization and introduction of new foods. Therefore, it is necessary to update the existing Cuban tables.

Aim:

Implement a procedure that makes it possible to process data on foods, coming from the literature, with the interest of developing reliable tables of nutritional composition of foods for ruminant animals.

Materials and Methods:

AliCuba software was used to generate tables of nutritional composition of livestock foods, based on the compilation and processing of research results published in various bibliographic sources. A procedure was established that regulates the stages to be followed, with the help of this software, to obtain reliable tables of the nutritional composition of foods. Univariate techniques were used to determine outlier data. Missing values were estimated through equations. Bootstrap resampling was applied with a percentile confidence interval. The procedure was validated with information regarding Megathyrsus maximus cv. Likoni (leaves and stems, fresh, western region, without specifying time of year or irrigation, with fertilization and 30 to 60 days old).

Results:

The procedure implemented following its different stages was adequate to identify and adjust database anomalies, such as data entry errors, incorrect units of measurement, analytical errors, misidentification of foods, etc., demonstrating its validity. to ensure the quality of the information.

Conclusion:

A procedure was implemented that allows perfecting the process of preparing nutritional composition tables of foods for ruminant animals.