Application of data mining techniques to predict the production of aflatoxin B1 in dry-cured ham

Belén Peromingo; Daniel Caballero; Alicia Rodríguez; Andrés Caro; Mar Rodríguez

doi:10.1016/j.foodcont.2019.106884

Application of data mining techniques to predict the production of aflatoxin B₁ in dry-cured ham

Belén Peromingo, Daniel Caballero, Alicia Rodríguez, Andrés Caro, Mar Rodríguez^*

^*Corresponding author for this work

Food Analytics and Biotechnology

3 Citations (Scopus)

Abstract

Dry-cured ham may be contaminated with aflatoxin B₁ (AFB₁) produced by Aspergillus spp. Temperature and water activity (a_w) are two key parameters that affect both ham ripening and AFB₁ production. The objective of this study was to predict AFB₁ production by Aspergillus parasiticus and Aspergillus flavus strains in conditions related to dry-cured ham ripening using data mining techniques. J48 decision tree, isotonic regression (IR), and multiple linear regression (MLR) were tested to (a) classify and predict AFB₁ concentration as a function of different days, temperatures and a_w values and (b) predict the beginning of AFB₁ production as a function of different temperatures and a_w values. For this, a model system based on a dry-cured ham-based medium was used. The percentage of correct classification was higher than 75%. R values to predict the concentration of AFB₁ when applying MLR were 0.81, being higher than those obtained after using IR. The models developed were validated with experimental data obtained after inoculating samples of dry-cured ham with two aflatoxigenic strains. The predicted AFB₁ concentration showed correlation coefficients ≥0.74 and prediction errors ≤0.38, confirming the feasibility of the prediction equations obtained. This information may help to make informed decisions to minimise the hazard posed by AFB₁ in dry-cured ham.

Original language	English
Article number	106884
Journal	Food Control
Volume	108
Number of pages	7
ISSN	0956-7135
DOIs	https://doi.org/10.1016/j.foodcont.2019.106884
Publication status	Published - 2020

Keywords

Aflatoxins
Aspergillus spp.
Data mining
Dry-cured ham
Prediction

Access to Document

10.1016/j.foodcont.2019.106884

Cite this

@article{2c421061a4774fd18a030115582b2404,

title = "Application of data mining techniques to predict the production of aflatoxin B1 in dry-cured ham",

abstract = "Dry-cured ham may be contaminated with aflatoxin B1 (AFB1) produced by Aspergillus spp. Temperature and water activity (aw) are two key parameters that affect both ham ripening and AFB1 production. The objective of this study was to predict AFB1 production by Aspergillus parasiticus and Aspergillus flavus strains in conditions related to dry-cured ham ripening using data mining techniques. J48 decision tree, isotonic regression (IR), and multiple linear regression (MLR) were tested to (a) classify and predict AFB1 concentration as a function of different days, temperatures and aw values and (b) predict the beginning of AFB1 production as a function of different temperatures and aw values. For this, a model system based on a dry-cured ham-based medium was used. The percentage of correct classification was higher than 75%. R values to predict the concentration of AFB1 when applying MLR were 0.81, being higher than those obtained after using IR. The models developed were validated with experimental data obtained after inoculating samples of dry-cured ham with two aflatoxigenic strains. The predicted AFB1 concentration showed correlation coefficients ≥0.74 and prediction errors ≤0.38, confirming the feasibility of the prediction equations obtained. This information may help to make informed decisions to minimise the hazard posed by AFB1 in dry-cured ham.",

keywords = "Aflatoxins, Aspergillus spp., Data mining, Dry-cured ham, Prediction",

author = "Bel{\'e}n Peromingo and Daniel Caballero and Alicia Rodr{\'i}guez and Andr{\'e}s Caro and Mar Rodr{\'i}guez",

year = "2020",

doi = "10.1016/j.foodcont.2019.106884",

language = "English",

volume = "108",

journal = "Food Control",

issn = "0956-7135",

publisher = "Pergamon Press",

}

TY - JOUR

T1 - Application of data mining techniques to predict the production of aflatoxin B1 in dry-cured ham

AU - Peromingo, Belén

AU - Caballero, Daniel

AU - Rodríguez, Alicia

AU - Caro, Andrés

AU - Rodríguez, Mar

PY - 2020

Y1 - 2020

N2 - Dry-cured ham may be contaminated with aflatoxin B1 (AFB1) produced by Aspergillus spp. Temperature and water activity (aw) are two key parameters that affect both ham ripening and AFB1 production. The objective of this study was to predict AFB1 production by Aspergillus parasiticus and Aspergillus flavus strains in conditions related to dry-cured ham ripening using data mining techniques. J48 decision tree, isotonic regression (IR), and multiple linear regression (MLR) were tested to (a) classify and predict AFB1 concentration as a function of different days, temperatures and aw values and (b) predict the beginning of AFB1 production as a function of different temperatures and aw values. For this, a model system based on a dry-cured ham-based medium was used. The percentage of correct classification was higher than 75%. R values to predict the concentration of AFB1 when applying MLR were 0.81, being higher than those obtained after using IR. The models developed were validated with experimental data obtained after inoculating samples of dry-cured ham with two aflatoxigenic strains. The predicted AFB1 concentration showed correlation coefficients ≥0.74 and prediction errors ≤0.38, confirming the feasibility of the prediction equations obtained. This information may help to make informed decisions to minimise the hazard posed by AFB1 in dry-cured ham.

AB - Dry-cured ham may be contaminated with aflatoxin B1 (AFB1) produced by Aspergillus spp. Temperature and water activity (aw) are two key parameters that affect both ham ripening and AFB1 production. The objective of this study was to predict AFB1 production by Aspergillus parasiticus and Aspergillus flavus strains in conditions related to dry-cured ham ripening using data mining techniques. J48 decision tree, isotonic regression (IR), and multiple linear regression (MLR) were tested to (a) classify and predict AFB1 concentration as a function of different days, temperatures and aw values and (b) predict the beginning of AFB1 production as a function of different temperatures and aw values. For this, a model system based on a dry-cured ham-based medium was used. The percentage of correct classification was higher than 75%. R values to predict the concentration of AFB1 when applying MLR were 0.81, being higher than those obtained after using IR. The models developed were validated with experimental data obtained after inoculating samples of dry-cured ham with two aflatoxigenic strains. The predicted AFB1 concentration showed correlation coefficients ≥0.74 and prediction errors ≤0.38, confirming the feasibility of the prediction equations obtained. This information may help to make informed decisions to minimise the hazard posed by AFB1 in dry-cured ham.

KW - Aflatoxins

KW - Aspergillus spp.

KW - Data mining

KW - Dry-cured ham

KW - Prediction

U2 - 10.1016/j.foodcont.2019.106884

DO - 10.1016/j.foodcont.2019.106884

M3 - Journal article

AN - SCOPUS:85072034048

SN - 0956-7135

VL - 108

JO - Food Control

JF - Food Control

M1 - 106884

ER -