**3. Results and Discussion**


Table 3 displays the composition (% of the FAs categories) of the nine initial oils from which the mixtures were made and the two commercial supplements of ω-3.


**Table 3.** Composition (%) of the initial oils with their standard deviation.

For each sample, three replications were performed. SFAs: saturated fatty acids, MUFAs: monounsaturated fatty acids, PUFAs: polyunsaturated fatty acids, ω-3: omega-3 fatty acids, ω-6: omega-6 fatty acids.

Oil samples show high variability in their FAs profiles. This suggests that the oils collected for this study might have different origins and could come from different kinds of fish, production methods or various types of processing industries. This sample variability highlights the importance of determining the lipid profile of fish oils, as the percentage of the different groups, especially ω-3, varies significantly between samples.

On the one hand, oils A, C and F and Supplement B showed the typical seawater fish oil composition regarding PUFAs, where most of the PUFAs come from ω-3 FAs [31–33]. In these samples, PUFAs represented between 30% and 46% of the total FAs of the oils. They had an elevated ω-3 content, which almost corresponded with all the PUFAs in the samples, and a lower content of ω-6. Considering their composition, these samples may come from a process where only seawater fish is involved, i.e., fish fillet processing [31,32].

On the other hand, samples B, D, E, G, H and I and supplement A had PUFAs content between 25% and 34%, which is also typical in fish oil [31]. However, these samples presented a higher content of ω-6 FAs than the previous set, in which ω-3 FAs were predominant. This is due to a high level of linoleic acid (18:2) (data not shown), which may have two explanations: on the one hand, it might be due to the fish species from which the oil was obtained, i.e., this PUFAs profile is characteristic of freshwater fish [34], which has a higher ω-6 content in comparison with seawater fish [32,33]. On the other hand, it might be due to the type of fish processing industry from which the samples originated. In the canning industry, by-products of fish oil are mixed with vegetable oils, such as sunflower oil, which is rich in ω-6 PUFAs (linoleic acid) [35,36].

### 3.1.2. Fatty Acid Profile of the Oil Mixtures

The results of the characterization and the statistics of the oil samples in the different sets of data used in the models are shown in Table 4.


**Table 4.** Results of the characterization of all the samples.

*n*: number of samples, SD: standard deviation, SFAs: saturated fatty acids, MUFAs: monounsaturated fatty acids, PUFAs: polyunsaturated fatty acids, ω-3: omega-3 fatty acids, ω-6: omega-6 fatty acids.

MUFAs constitute the majority group in most cases representing: in the 1st step, 41.4% of total FAs composition on the calibration set and, in the 2nd step, the 40.1% of the calibration set and the 43.7% of the external validation set. However, PUFAs are the majority group in the validation set of the 1st step, with a percentage of 38.6%. On the contrary, ω-6 is the least common group in the four sets of samples, with percentages of 10.3% and 8.6% in the calibration and validation set of the first step, respectively, and 9.8% and 12.1% in the calibration and external validation set of the second step, respectively.

### *3.2. Performance of the PLSR Models of the Target Oils*
