Kevin Goodner

The dangers of creating false classifications due to noise in electronic nose and similar multivariate analyses

Abstract Randomly generated data with the error limits of 1–10% along with experimental data was employed to demonstrate the dangers of over-fitting data which creates artificial differentiation. Analysis of variance (ANOVA), principal components analysis (PCA), and discriminant function analysis (DFA) were employed for the data analysis. In cases, where the ratio of samples to variables (features) falls below six, single class systems containing only random noise and random groupings can be misclassified into more than a single group when the discriminate techniques are employed. The smaller the group size, the more erroneous classifications are made. Larger sample sizes minimize the random noise and allow the true differences to show. A minimum number of variable (features) should be employed with developing classification models to avoid over-fitting data. The ratio of data points to variables should be at least six to avoid over-fitting classification errors with validation of the model using data points not used in generating the model.

Distribution of aroma volatile compounds in tangerine hybrids and proposed inheritance

BACKGROUND With the desirable combination of sugars and acids, volatile compounds contribute to the essential organoleptic attributes of citrus. This study evaluated the aroma volatiles of 20 tangerine hybrids of the University of Florida breeding program. Volatiles were sampled from hand-squeezed juice by headspace solid-phase microextraction (SPME), and analyzed by gas chromatography-mass spectrometry. Principal component analysis (PCA) and cluster analysis (CA) were used to find similarities among samples due to volatile composition with effect of genetic background. RESULTS In total, 203 volatiles were detected in all samples. Volatiles in lower amounts were widely distributed among samples and were classified mainly as terpene hydrocarbons and oxygenated compounds, such as aldehydes, esters, alcohols and ketones. PCA, based on relative peak areas (content) clearly separated the samples higher in volatile content, mainly those with sweet orange genetic contributions in their background. CA, based on volatile presence/absence, grouped samples into five clusters, each showing distinctive volatile profiles. CONCLUSION The genetic background of tangerine hybrids affected volatile composition and content of samples. In general, tangerines were characterized by fewer volatiles (in both quality and quantity) and more aldehydes, and hybrids with sweet orange in their background had more sesquiterpenes and esters, which would likely affect their aroma.

Deaeration and pasteurization effects on the orange juice aromatic fraction

Abstract A comparative study between the aromatic profile in fresh orange juice versus deaerated and pasteurized juices, respectively, was conducted in order to understand the evolution of volatile components after deaeration and pasteurization processes. The aromatic fractions isolated by simultaneous distillation and extraction were analysed by capillary gas chromatography–mass spectrometry. At the qualitative level all the volatile components in fresh orange juice were also found in the counterparts after deaeration and pasteurization processes. According to statistical analyses, significant losses in concentration of volatile components occurred during the deaeration process, while there were no statistically significant differences determined among concentrations of volatile components in deaerated and pasteurized juices. These results show that during the industrial processing of orange juice the biggest losses in the concentration of volatile components occurs during deaeration. The pasteurization process does not change the analytical composition of deaerated orange juice in a significant way for any of the 42 quantitated compounds.

Comparison of grapefruit hybrid fruit with parent fruit based on composition of volatile components

Two new grapefruit hybrids and their parent cultivars were evaluated by headspace gas chromatographic analysis of fresh juice from the fruit. Quantities of 39 volatile components were determined, and the variation compared to the parent fruit for each component monitored was tabulated. The relatively low amounts of some volatile components in hybrid 2 fruit support the observation that the flavor of juice from this hybrid fruit is milder than that of grapefruit. Multivariate analysis using principal component and discriminant analyses afforded more information on the similarities and differences of fruit of each hybrid to its parents’ fruits and to that of the other hybrid. Discriminant analysis showed the hybrid fruit to be similar to each other and dissimilar to the parent fruit based on the volatile components monitored. One hybrid is useful for cross-breeding to produce other new grapefruit-like hybrids, and the other has potential as a new commercial grapefruit-type fruit since it matures very early and has an excellent grapefruit flavor.

Quantification of β-damascenone in orange juice using headspace standard addition SPME with selected ion GC-MS

An analytical procedure to quantify concentrations of β-damascenone in orange juice was developed using static headspace SPME with selected ion GC-MS at m/z 190. Standard additions at four levels of β-damascenone (0.5 to 2 μg L−1) were used to establish the regression equation used to quantify juice samples. Reproducibility in terms of relative standard deviation values at each level ranged from 1% to 15%. Orange juice β-damascenone concentrations increased with increased heat treatment from 0.15 μg L−1 in the raw juice to 0.3 μg L−1 in juice which had been heated to 90 °C for 25 s. Levels of β-damascenone were determine in three types of commercial orange juice which have different levels of heat treatment. β-damascenone concentrations ranged from 0.122 to 0.281 μg L−1 in not from concentrate, NFC, juice (heated once), 0.117 to 0.445 μg L−1 in frozen concentrated orange juice, FCOJ, and 0.221 to 0.690 μg L−1 in reconstituted from concentrate, RFC, (heated to remove water and heated again after water restored). Although levels of β-damascenone increase with heat treatment, the wide range of β-damascenone concentrations observed would make it inappropriate to use as a single indicator of heat treatment. This is the first time β-damascenone has been quantified by headspace SPME in different types of orange juice samples.