The hypothesis in this study is that methyl esters from wistar rat animal fat can be analyzed using the Gas Chromatography Mass Spectrometry (GCMS) method. The methyl ester data combined with chemometrics is able to classify types of fat.

Samples in the form of pork, beef, dog, goat, wild boar, chicken were obtained from the traditional market, Wistar white rats were obtained from other researchers' carcasses, the fat was taken. Materials used n-hexane (technical), NaOCH₃ 0.2 N solution (E-Merck, pro-analysis quality), BF₃ solution (E-Merck, pro-analysis quality), saturated NaCl and anhydrous Na₂SO₄ (E-Merck, pro-analysis quality) (Rohman and Che Man, 2011; Kumar et al., 2014).

GC-MS distributor from Ditek Jaya, Merck Shimadzu, Japan, type GCMS-QP2010 SE. The column used was Rtx-5ms, DB1-MS Restech, 30 m x 0.25 mm ID, 0.25 μm, stationary phase of polymethyl xiloxan, injector temperature of 230 °C, column temperature of 70 °C and increased to 300 °C with an increase of 10 °C.min^-1, and flow rate of 1.15 mL.min^-1. The mobile phase of Helium gas. MS Detector Electron Multifier Detector (EMD) 70 MeV. The mass spectrum was compared to the WILLEY147 & NIST47 library found in the GC-MS software. The methyl ester was injected as much as μL into the GC column in a manner of autosampler.

Rendering extraction to obtain rat fat. The advantages of this method are that there are many extract yields, easy and inexpensive processing because it does not involve chemicals (Rohman et al., 2016). The yield of 16.36%. The yields are influenced by the intake of food from the rats themselves, body fat taken, and the way the fat is extracted (Lobb and Chow, 2007). The obtained lipid fraction was then carried out by the esterification process. The esterification reaction aims to convert the fatty acids into their methyl ester forms. The esterification process is carried out using BF₃ as a catalyst. BF₃ is an acidic compound (Purbasari and Silviana, 2008). Figure 1 is a process of the esterification mechanism and the product of methyl ester is obtained.

Analysis of methyl esters from Wistar rat derivatization using gas chromatography-mass spectrometry (GC-MS) method. This instrument is a combination of gas chromatography with a mass spectrometer detector. Gas chromatography is used for the separation of fatty acid content in the form of methyl esters. In addition to the retention time (tR) of the separation results in gas chromatography, there is similarity index (SI) information to determine the proximity of the chemical structure of the type of fatty acids. The result of SI >90 shows the similarity of mass ion overflow to the target spectra/fat sample.

Table 1 presents retention time (tR), % peak area, SI, molecular weight (MW) and estimated compounds and identification of white rat fat (Guntarti and Amidin, 2018). The results of GC-MS analysis in Table I show the results of SI values >90 except oleic acid which is 88% and linoleic acid 86%. This shows that the target fatty acid type is suitable or similar to the comparison spectra. The fatty acid compound with a tR of 25.93 minutes and a SI value of 96% was similar to the comparison compound with the formula C₁₇H₃₆O₂ with m/z 296. The fatty acid was in the form of its methyl ester. Whereas if in the form of fatty acids, the compound formula is C₁₈H₃₃O₂.

Fat is an unstable component in the presence of light. The results of the analysis with GC-MS showed that oleic acid was the highest constituent component of fatty acids in rat fat with a percentage of 40.48%, followed by linoleic acid 30.14%, palmitic acid 19.08%, stearic acid 2.55%, palmitoleic acid 0.73%, and myristic acid 0.15%. The line of the types of fatty acids in Wistar rat fat is presented in Figure 2. unsaturated fatty acid with one double bond. When viewed from unsaturated bonds, white rat fat contains many types of unsaturated fatty acids, namely palmitoleic acid (0.73%), linoleic acid (30.14%), and oleic acid (40.48%). Whereas saturated fatty acids are myristic acid (0.15%), palmitic acid (19.08%), and stearic acid (2.55%). If looked at the percentage of the content, then more unsaturated fatty acids is equal to 71.35%. Saturated fatty acids of 21.78% and 6.87% are ingredients other than methyl esters. Larger amount of unsaturated fatty acid content will affect the physical form of fat at room temperature and the stability of fat. Figure 3 presents a line of total fatty acid, saturated fatty acid, unsaturated fatty acid, and ingredients other content in Wistar rat.

Besides Wistar rat fat, other animals’ fat used were: dog, beef, pork, chick, and goat. Fat retrieval is the same as done in white fat retrieval, which is by rendering with an oven at a temperature of 90 °C – 100 °C, for 30 – 60 minutes. The fat obtained is esterified to form its methyl ester with NaOCH₃ and BF₃ which are then injected into the GC-MS system. The results of analysis of dog, beef, pork, chick, goat, and wild boar fat are presented in Table II and Figure 4.

Based on Table II it can be seen that the fat of white rat and pork containing linoleic content (30.14%), and pork fat (21.49%). The highest oleic acid is in the content of pork (55.66%). Goat fat has the highest type of saturated fat, palmitate (23.55%) and stearic (47.13%).

In the results of previous studies (Hermanto, Muawanah and Harahap, 2008), that pork and chicken fat contain margaric acid. Margaric acid content (C17: 0) in pork is 0.5%, and in chick 1.74% (Hermanto, Muawanah and Harahap, 2008). Guntarti (2018) research results: beef fat has a high stearic acid (35.03%), while oleic acid is 14.90%. The results of this study, beef fat has a high oleic acid content (52.29%), while stearic acid is 12.59%. Except for goat fat, all animals contain the highest oleic acid; while goat fat, the highest is stearic acid. Figure 4 presents the content of saturated and unsaturated fat, and the total amount of fatty acids in various animals. Based on Figure 4, the saturated fatty acid content is high in goat fat (71.16%), wild boar fat (36.27%), beef fat (35.16%), dog fat (34.33%), lard/pork (27.78%), and the smallest is in chick fat (20.55%). The highest unsaturated fat content is in chick fat (73.2%), rat fat (71.35%), lard / pork fat (55.66%), dog fat (53.93%), wild boar fat (45.24%), and the smallest is in goat fat (19.19%). While the highest amount of total fat is in chick, followed by white rat.

PCA data interpretation is done by reducing data, in which the number of variables in a matrix is reduced to produce new variables while maintaining the information held by the data. The new variables generated are scores or main components (Rohman and Man, 2012). PCA aims to group variables that are correlated with each other and replace them with new groups called main components (principal component) (Coltro et al., 2005). PCA simplifies data by reducing a number of variables to a smaller number of orthogonal variables. This needs a correlation between variables. Although PCA reduces the number of initial variables, PCA retains variability and initial information. PCA also helps provide pattern visualization and correlation analysis (Miller and Miller, 2010). PCA plot scores are presented in Figure 5. The results of replication greatly affect the location of the quadrants obtained, further showing that there are similarities in the physical chemical properties of the fatty acid content. Wistar rat fat is located between chick fat and lard. The results of replication measurements affect the proximity position of the grouping of animal fat.

The results of PCA analysis using Minitab resulted in 8 PCs presented in Figure 6. Each PC displays eigenvalue, proportion, and cumulative values. Eigenvalue variations can explain the data on each PC and show how much influence a variable on the formation of the characteristics of a matrix (Miller and Miller, 2010). In Table 3, PC1 with eigenvalue 3.8699 is able to describe 48.4% of the total original data variables while PC2 with eigenvalue 2.3938 is able to describe 21.90% of the total original variables, PC3 with eigenvalue 1.0508 is able to describe 13.10%. Thus, 3 PCs described the illustration data for discriminant analysis of 83.40%.