Multivariate Data Analysis Techniques Research Articles

Changes in Black Truffle (Tuber melanosporum) Aroma during Storage under Different Conditions.

The enticing aroma of truffles is a key factor for their culinary value. Although all truffle species tend to be pricy, the most intensely aromatic species are the most sought after. Research into the aroma of truffles encompasses various disciplines including chemistry, biology, and sensory science. This study focusses on the chemical composition of the aroma of black truffles (Tuber melanosporum) and the changes occurring under different storage conditions. For this, truffle samples were stored under different treatments, at different temperatures, and measured over a total storage time of 12 days. Measurements of the truffle aroma profiles were taken with SPME/GC-MS at regular intervals. To handle the ample data collected, a systematic approach utilizing multivariate data analysis techniques was taken. This approach led to a vast amount of data which we made publicly available for future exploration. Results reveal the complexity of aroma changes, with 695 compounds identified, highlighting the need for a comprehensive understanding. Principal component analyses offer initial insights into truffle composition, while individual compounds may serve as markers for age (formic acid, 1-methylpropyl ester), freshness (2-Methyl-1-propanal; 1-(methylthio)-propane), freezing (tetrahydrofuran), salt treatment (1-chloropentane), or heat exposure (4-hydroxy-3-methyl-2-butanone). This research suggests that heat treatment or salt contact significantly affects truffle aroma, while freezing and cutting have less pronounced effects in comparison. The enrichment of compounds showing significant changes during storage was investigated with a metabolomic pathway analysis. The involvement of some of the enriched compounds on the pyruvate/glycolysis and sulfur pathways was shown.

A Review on the Recent Advancements and Artificial Intelligence in Tablet Technology.

Tablet formulation could be revolutionized by the integration of modern technology and established pharmaceutical sciences. The pharmaceutical sector can develop tablet formulations that are not only more efficient and stable but also patient-friendly by utilizing artificial intelligence (AI), machine learning (ML), and materials science. The primary objective of this review is to explore the advancements in tablet technology, focusing on the integration of modern technologies like artificial intelligence (AI), machine learning (ML), and materials science to enhance the efficiency, cost-effectiveness, and quality of tablet formulation processes. This review delves into the utilization of AI and ML techniques within pharmaceutical research and development. The review also discusses various ML methodologies employed, including artificial neural networks, an ensemble of regression trees, support vector machines, and multivariate data analysis techniques. Recent studies showcased in this review demonstrate the feasibility and effectiveness of ML approaches in pharmaceutical research. The application of AI and ML in pharmaceutical research has shown promising results, offering a potential avenue for significant improvements in the product development process. The integration of nanotechnology, AI, ML, and materials science with traditional pharmaceutical sciences presents a remarkable opportunity for enhancing tablet formulation processes. This review collectively underscores the transformative role that AI and ML can play in advancing pharmaceutical research and development, ultimately leading to more efficient, reliable and patient-centric tablet formulations.

SOCIAL HEALTH INSURANCE ADMINISTRATION BODY PARTICIPANTS’ SATISFACTION ON DECISION TO USE MOBILE APPLICATION

Mobile JKN is an application developed by Social Health Insurance Administration Body to provide easy access to information about the National Healthcare Insurance Program (JKN Program). The number of participants in the Social Security Agency on Health (BPJS Health) has almost reached 250 million people. The government must manage public services so Social Security Agency on Health participants can access health facilities conveniently. The study aims to analyze the Social Security Agency on Health participants’ satisfaction with the decision to use the Mobile JKN application using the Expectancy Disconfirmation Model (EDM). EDM described how perceived expectations and performance influence public satisfaction. The study method is quantitative, with the study unit being Social Security Agency on Health participants and locus in Bogor. Structural Equation Modeling (SEM), a multivariate data analysis technique, was used in this study to examine the data and assess the intricate relationships between the constructs and indicators. The data processing technique uses SmartPLS. The finding shows that using the mobile JKN application has positively impacted user satisfaction. The Social Security Agency on Health participants were satisfied after utilizing the Mobile JKN application to access healthcare services.

Metabolomics combined with chemometric analysis to identify α-glucosidase inhibitors in Phaleria macrocarpa fruit extracts and its molecular docking simulation

Phaleria macrocarpa is a medicinal plant widely available in Malaysia. The plant parts have been traditionally used as an antidiabetic remedy. This study aimed to identify the putative inhibitors of the α-glucosidase enzyme from P. macrocarpa using a gas chromatography-mass spectrometry (GC–MS)-based metabolomics approach and further subjected them to in silico molecular docking analysis to elucidate the possible mechanism of action. This study assessed the inhibitory potential of P. macrocarpa fruit extracts (aqueous, 20 %, 40 %, 60 %, 80 %, and 100 % ethanol) against the α-glucosidase enzyme using GC–MS and chemometric analysis. Orthogonal partial least squares (OPLS) combined with GC–MS analysis were applied to correlate the inhibition of enzyme activity to the metabolites profile of P. macrocarpa. All the potential inhibitors identified were further docked to the yeast (Saccharomyces cerevisiae) protein crystal structure (PDB3A4A). The generated score scatter plot of OPLS showed a recognizable separation of all the extracts into six different clusters. GC–MS, incorporated with multivariate data analysis techniques, was used to identify significant chemical markers. Methyl α-d-glucopyranoside, squalene, palmitic acid, myo-inositol and isoquinoline metabolites were identified as putative inhibitors against α-glucosidase activity from P. macrocarpa. In conclusion, the GC–MS-based metabolomics approach identified potential chemical markers of P. macrocarpa that could be utilized in the development of herbal based medicine.

Integrated Electrochemical Aptasensor Array toward Monitoring Anticancer Drugs in Sweat.

In the realm of clinical practice, the concurrent utilization of anticancer medications can enhance their overall therapeutic efficacy. However, it is crucial to acknowledge that the interactions among these anticancer drugs can potentially yield detrimental consequences on their intended outcomes. Consequently, the assessment of both anticancer potency and potential toxic side effects is greatly refined when multiple anticancer drugs are simultaneously detected and evaluated. Here, we designed a wearable electrochemical aptasensor array for monitoring multiple anticancer drugs in sweat. The integrated sensor array consists of three working electrodes modified with three different aptamers (Apt1, Apt2, and Apt3), a Au counter electrode, and a Ag/AgCl reference electrode. Molecular docking simulations were performed to show the binding affinities between three anticancer drugs and their corresponding aptamers. Various eigenvalues were derived from the square-wave voltammetry electrochemical signals, and these data sets were subjected to rigorous analysis through multivariate data analysis techniques. This analytical approach demonstrated exceptional performance by achieving flawless 100% accuracy in the precise identification of nine anticancer drugs consistently at uniform concentrations. Furthermore, the integrated wearable sensor array exhibited impressive capabilities, correctly recognizing all nine anticancer drugs with 100% accuracy and successfully distinguishing between these drugs in artificial sweat samples. The proposed sensor array presents good stability for 15 days. Flexibility tests showed stable device performance after 500 twisting cycles. This innovative wearable sensing array represents a novel approach for achieving real-time monitoring and precise adjustment of drug dosages. It offers invaluable insights for tailoring the treatment of anticancer drugs to individual patients, predicting both drug efficacy and potential adverse reactions within the field of clinical medicine.

Urban Indian Environment in the Context of a Pandemic

The spread of the Coronavirus disease 2019 (COVID-19) has impacted human life severely since November 2019. The urban centers in the world, especially, were highly affected by the diseases. Several socioeconomic and environmental factors probably enhanced the spread of the pandemic and consequent mortality. Many studies examining environmental factors, such as air quality, in urban centers indicate the roles of those factors in the spread of diseases and consequent mortality. However, other socioeconomic factors that directly or indirectly elevate the mass death of people are seldom studied. The present study explores the socioeconomic factors and air quality influencing COVID-19 deaths in urban India. We randomly selected 19 Indian cities and collected each city’s socioeconomic and air quality data from reliable and open sources. The data were analyzed using multivariate data analysis techniques using R statistics. The results showed significant positive relationships, population, and total area of the urban centers.

Plasma Lipidomic Profiling Using Mass Spectrometry for Multiple Sclerosis Diagnosis and Disease Activity Stratification (LipidMS).

This investigation explores the potential of plasma lipidomic signatures for aiding in the diagnosis of Multiple Sclerosis (MS) and evaluating the clinical course and disease activity of diseased patients. Plasma samples from 60 patients with MS (PwMS) were clinically stratified to either a relapsing-remitting (RRMS) or a chronic progressive MS course and 60 age-matched controls were analyzed using state-of-the-art direct infusion quantitative shotgun lipidomics. To account for potential confounders, data were filtered for age and BMI correlations. The statistical analysis employed supervised and unsupervised multivariate data analysis techniques, including a principal component analysis (PCA), a partial least squares discriminant analysis (oPLS-DA) and a random forest (RF). To determine whether the significant absolute differences in the lipid subspecies have a relevant effect on the overall composition of the respective lipid classes, we introduce a class composition visualization (CCV). We identified 670 lipids across 16 classes. PwMS showed a significant increase in diacylglycerols (DAG), with DAG 16:0;0_18:1;0 being proven to be the lipid with the highest predictive ability for MS as determined by RF. The alterations in the phosphatidylethanolamines (PE) were mainly linked to RRMS while the alterations in the ether-bound PEs (PE O-) were found in chronic progressive MS. The amount of CE species was reduced in the CPMS cohort whereas TAG species were reduced in the RRMS patients, both lipid classes being relevant in lipid storage. Combining the above mentioned data analyses, distinct lipidomic signatures were isolated and shown to be correlated with clinical phenotypes. Our study suggests that specific plasma lipid profiles are not merely associated with the diagnosis of MS but instead point toward distinct clinical features in the individual patient paving the way for personalized therapy and an enhanced understanding of MS pathology.

An introductory review on the application of principal component analysis in the data exploration of the chemical analysis of food samples.

Principal component analysis (PCA) is currently one of the most used multivariate data analysis techniques for evaluating information from food analysis. In this review, a brief introduction to the theoretical principles that underlie PCA will be given, in addition to presenting the most commonly used computer programs. An example from the literature was discussed to illustrate the use of this chemometric tool and interpretation of graphs and parameters obtained. A list of recently published articles will also be presented, in order to show the applicability and potential of the technique in the food analysis field.

Surface-Enhanced Raman Spectroscopy for Evaluation of Antibacterial Activity of the N-Heterocyclic Carbene Selenium Complex and Its Ligand Against Pathogenic Bacterial Strains Using Multivariate Data Analysis Techniques

Surface-enhanced Raman spectroscopy has been employed to analyze the antibacterial potential of an N-heterocyclic carbene selenium complex (NHC Se complex), its ligand, and commercial derivative ornidazole. Surface-enhanced Raman spectroscopy (SERS) is an analytical method that enhances Raman signals through the interaction of molecules with nanostructured metallic surfaces. Two pathogenic bacteria, including gram positive Staphylococcus aureus (S. aureus) and gram-negative Escherichia Coli (E. coli), were used to characterize the antibacterial activity of the complex and ligand. After the application of selenium complex and its ligand, notable spectral changes were observed in mean SERS spectra of both bacteria treated with these antibacterial agents. SERS spectral features of bacterial cell metabolites at 657 cm−1, 728 cm−1, 1091 cm−1 and 1462 cm−1 decreased after treatment by the complex and its ligand. However, the intensity of features at 1410 cm−1 and 1226 cm−1 were enhanced after the application of complex and its ligand. SERS was an efficient tool in demonstrating the trend of the spectral data of bacteria treated with different concentrations of antibiotics. Three multivariate data analysis techniques, including principal component analysis (PCA), partial least square discriminant analysis (PLS-DA) and partial least square regression (PLSR), discriminated the spectra of those treated and untreated with the antibacterial agents.

Identification of provenance of Basmati rice grown in different regions of Punjab through multivariate analysis

High-priced Basmati rice is vulnerable to deliberate mislabeling to increase profits. This type of fraud may lower consumers’ confidence as inferior products can affect brand reputation. To address this problem, there is a need to devise a method that can efficiently distinguish Basmati rice grown in regions that are famous versus the regions that are not suitable for their production. Therefore, in this investigation, thirty-six samples of Basmati rice were collected from two zones of Punjab province (one known for Basmati rice) of Pakistan which is the major producer of Basmati rice. The elemental composition of rice samples was assessed using inductively coupled plasma-optical emission spectrometry and an organic elemental analyzer, whereas data on δ13C was acquired using isotopic ratio-mass spectrometry. Regional clustering of samples based on their respective cultivation zones was observed using multivariate data analysis techniques. Partial least squares-discriminant analysis was found to be effective in grouping rice samples from the different locations and identifying unknown samples belonging to these two regions. Further recommendations are presented to develop a better model for tracing the origin of unidentified rice samples.

Quantitative analysis of cephalexin in solid dosage form by Raman spectroscopy and chemometric tools

Objective To use Raman Spectroscopy for qualitative and quantitative evaluation of pharmaceutical formulations of active pharmaceutical ingredient (API) of Cephalexin. Significance Raman Spectroscopy is a noninvasive, nondestructive, reliable and rapid detection technique used for various pharmaceutical drugs quantification. The present study explores the potential of Raman Spectroscopy for quantitative analysis of pharmaceutical drugs. Method For qualitative and quantitative analysis of Cephalexin API, various standard samples containing less and more concentration of API than commercial tablet was prepared. To study spectral differences, the mean plot of all the samples was prepared. For qualitative analysis, Principal Component Analysis (PCA) and for quantitative analysis Partial Least Square Regression analysis (PLSR) was used. Both of these are Multivariate data analysis techniques and give reliable results as published in previous literature. Results PCA model distinguished all the Raman Spectral data related to the various Cephalexin solid dosage formulations whereas the PLSR model was used to calculate the concentration of different unknown formulations. For the PLSR model, RMSEC and RMSEP were determined to be 3.3953 and 3.8972, respectively. The prediction efficiency of this built PLSR model was found to be very good with a goodness of the model value (R2) of 0.98. The PLSR model also predicted the concentrations of Cephalexin formulations in the blind or unknown sample. Conclusion These findings demonstrate that the Raman spectroscopy coupled to PLSR analysis could be regarded as a fast and effectively reliable tool for quantitative analysis of pharmaceutical drugs.

Deciphering the mysteries of Aconitum pendulum: Unique identification of various processed products and characteristic chemical markers

Aconitum pendulum, a member of the Aconitum genus within the Ranunculaceae family, is known for its remarkable anti-inflammatory, anti-tumor, and analgesic properties. However, the application of Aconitum pendulum in traditional medicine is often restricted by its inherent toxicity, necessitating specific processing methods to ensure safe usage. Tibetan medicine utilizes distinctive processing techniques such as the barley wine, Hezitang, and zanba frying systems to render the herb safe for medicinal use. Despite the known pharmacological importance, scant research exists on how these processing methods affect the phytochemical profile of Aconitum pendulum, indicating a necessity for comparative analysis of the resulting concoctions. In this study, we employed 1H nuclear magnetic resonance (1H NMR) spectroscopy and ultra-performance liquid chromatography coupled with quadrupole time of flight mass spectrometry (UPLC-Q-TOF-MS) to explore the metabolomic variations among different processed products of Aconitum pendulum. Multivariate data analysis techniques, including pattern recognition, allowed us to investigate the intricacies of the compound profiles. To discover signature chemical markers that can differentiate the various processed products, we conducted paired t-tests and generated heat maps highlighting the detected differences. Our integrated analysis of the 1H NMR and UPLC-Q-TOF-MS data revealed distinct chemical signatures unique to each method of Aconitum pendulum processing. By employing chemometric analysis with variable importance in projection (VIP) scores greater than 1.0, and coupled with the statistical rigor of paired t-tests (P < 0.05), we successfully identified significant chemical markers in various processed products, including ferulic acid (FA), caffeic acid (CA), 3-acetylaconitine (AAC) and benzoylaconitine (BAC). Our findings provide a rapid and robust analysis of the chemical constituencies within raw and processed Aconitum pendulum. The discovery of potential chemical markers through an array of advanced pattern recognition techniques offers a substantial contribution to the pharmaceutical analysis. The results of this research pave the way for future comparative studies and facilitate the recognition of distinctively processed Aconitum pendulum products in traditional medicine.

Exploring the antiviral activity of α-ketoamides compounds through electronic structure calculations: a structure-activity relationship study

In recent years, the scientific community has worked intensively in the search and development of new drugs to suppress viral infections, such as COVID-19. In fact, a number of active compounds have been tested; however, the absence of significant structure-activity relationships hinders the production of optimized drugs. In this study, molecular modeling techniques were employed to investigate the electronic, structural and chemical reactivity properties of a set α-ketoamides whose antiviral activities have been reported in the literature, aiming to propose new promising derivatives. The local reactivity of the compounds was evaluated via condensed-to-atoms Fukui indexes and molecular electrostatic potential. Multivariate data analysis and random forests machine learning techniques were employed to correlate the antiviral properties and electronic and structural descriptors and identify relevant variables. A series of new derivatives were then proposed and evaluated via density functional theory-based calculations, and docking/molecular dynamics with the target protein of the virus. The results suggest that active derivatives present reduced reactivity towards electrophilic agents on the central core of the molecules and high reactivity on R1 ligands. Derivatives with higher predicted antiviral activities were proposed based on simple electronic descriptors, and their efficacies are reinforced by docking and molecular dynamics simulations. Communicated by Ramaswamy H. Sarma

Qualitative and Quantitative Analysis of Azithromycin as Solid Dosage by Raman Spectroscopy.

Active pharmaceutical ingredients (APIs) and excipients are main drug constituents that ought to be identified qualitatively and quantitatively. Raman spectroscopy is aimed to be an efficient technique for pharmaceutical analysis in solid dosage forms. This technique can successfully be used in terms of qualitative and quantitative analysis of pharmaceutical drugs, their APIs, and excipients. In the proposed research, Raman spectroscopy has been employed to quantify Azithromycin based on its distinctive Raman spectral features by using commercially prepared formulations with altered API concentrations and excipients as well. Along with Raman spectroscopy, principal component analysis and partial least squares regression (PLSR), two multivariate data analysis techniques have been used for the identification and quantification of the API. For PLSR, goodness of fit of the model (R2) was found to be 0.99, whereas root mean square error of calibration was 0.46 and root mean square error of prediction was 2.42, which represent the performance of the model. This study highlights the efficiency of Raman spectroscopy in the field of pharmaceutics by preparing pharmaceutical formulations of any drug to quantify their API and excipients to compensate for the commercially prepared concentrations.

Effect of business culture on organization commitment of banking sector employees

There has been a significant shift in the employment landscape in Vietnam's major cities, particularly in the banking sector after the COVID-19 outbreak. Many banks have to restructure their jobs and employees, thus impacting their cost management and operations. Banks need to take into account how to maintain the organizational commitment of employees in the long-term. This article investigates the effect of business culture on the organizational commitment of employees in the banking industry in Vietnam. A questionnaire survey is undertaken to collect information from 493 employees of the 31 top banks in order to accomplish this goal. The database was examined with the statistical program SPSS, by applying multivariate data analysis techniques to develop hypotheses. The findings showed that there are six factors that influence employees' commitment to commercial banks: (1) internal communication, (2) teamwork, (3) training and development, (4) reward and recognition, (5) consensus, and (6) core values. According to the findings of this study, organizational culture has a positive impact on employee commitment in the banking sector. The research thereby suggested some recommendations to build and adjust organizational culture to enhance the commitment of employees in commercial banks.

Effects of temperature-humidity index on blood metabolites of German dairy cows and their female calves

Heat stress (HS) impairs productivity, health, and welfare in dairy cows, and additionally causes metabolic changes. Hence, specific metabolites could be used as HS biomarkers. Consequently, the aim of the present study was to compare blood metabolite concentrations of German Holstein dairy cows and of their female calves suffering from high temperature-humidity index (THI) during late gestation (cows) or during their first week of life (calves) or not. According to the mean daily THI (mTHI) at the day before blood sampling, animals were classified into 2 groups: high mTHI ≥60 (hmTHI) and low mTHI <60 (lmTHI). To perform a standard cross-sectional 2-group study, cow groups (n = 48) and calf groups (n = 47) were compared separately. Differences in metabolite concentrations between hmTHI and lmTHI animals were inferred based on a targeted metabolomics approach. In the first step, processed metabolomics data were evaluated by multivariate data analysis techniques, and were visualized using the web-based platform MetaboAnalyst V5.0. The most important metabolites with pronounced differences between groups were further analyzed in a second step using linear mixed models. We identified 9 thermally sensitive metabolites for the cows [dodecanedioic acid; 3-indolepropionic acid; sarcosine; triglycerides (14:0_34:0), (16:0_38:7), (18:0_32:1), and (18:0_36:2); phosphatidylcholine aa C38:1; and lysophosphatidylcholine a C20:3] and for the calves [phosphatidylcholines aa C38:1, ae C38:3, ae C36:0, and ae C36:2; cholesteryl esters (17:1) and (20:3); sphingomyelins C18:0 and C18:1; and p-cresol sulfate], most of them related to lipid metabolism. Apart from 2 metabolites (3-indolepropionic acid and sarcosine) in cows, the metabolite plasma concentrations were lower in hmTHI than in lmTHI groups. In our heat-stressed dry cows, results indicate an altered lipid metabolism compared with lactating heat-stressed cows, due to the missing antilipolytic effect of HS. The results also indicate alterations in lipid metabolism of calves due to high mTHI in the first week of life. From a cross-generation perspective, high mTHI directly before calving seems to reduce colostrum quality, with detrimental effects on metabolite concentrations in offspring.

Computational algorithm for search and detection of metabolites in CDF(GCxGCxMS) files

Metabolomics is the most novel of the omic sciences, which is responsible for the study and comparison of metabolites. In this study, gas chromatography mass spectrometry (GC-MS) and two-dimensional gas chromatography time-of-flight mass spectrometry (GCxGC-TOFMS) were used for the identification of metabolites found in the blood, which, throw their results in CDF files for further analysis. Metabolic profiles of healthy patients were analyzed using multivariate data analysis techniques, which showed clear differences between the metabolites found, and compared with the universal Golm Metabolome Database (GMD), allowing them to be identified with a high percentage of coincidence. The results found reinforce the advantages of GCxGC-TOFMS and the role of metabolomics in the characterization of metabolites useful for different areas of biology.

Assessment of E-Senses Performance through Machine Learning Models for Colombian Herbal Teas Classification

This paper describes different E-Senses systems, such as Electronic Nose, Electronic Tongue, and Electronic Eyes, which were used to build several machine learning models and assess their performance in classifying a variety of Colombian herbal tea brands such as Albahaca, Frutos Verdes, Jaibel, Toronjil, and Toute. To do this, a set of Colombian herbal tea samples were previously acquired from the instruments and processed through multivariate data analysis techniques (principal component analysis and linear discriminant analysis) to feed the support vector machine, K-nearest neighbors, decision trees, naive Bayes, and random forests algorithms. The results of the E-Senses were validated using HS-SPME-GC-MS analysis. The best machine learning models from the different classification methods reached a 100% success rate in classifying the samples. The proposal of this study was to enhance the classification of Colombian herbal teas using three sensory perception systems. This was achieved by consolidating the data obtained from the collected samples.

Serum NMR metabolomics analysis of human metastatic colorectal cancer: biomarkers and pathway analysis.

We describe the use of NMR metabolomics to analyse blood serum samples from healthy individuals (n = 26) and those with metastatic colorectal cancer (CRC, n = 57). The assessment, employing both linear and non-linear multivariate data analysis techniques, revealed specific metabolite changes associated with metastatic CRC, including increased levels of lactate, glutamate, and pyruvate, and decreased levels of certain amino acids and total fatty acids. Biomarker ratios such as glutamate-to-glutamine and pyruvate-to-alanine were also found to be related to CRC. The study also found that glutamate was linked to progression-free survival and that both glutamate and 3-hydroxybutyrate were risk factors for metastatic CRC. Additionally, GC-FID was utilized to analyse the fatty acid profile and pathway analysis was performed on the profiled metabolites to understand the metabolic processes involved in CRC. A correlation was also found between the presence of certain metabolites in the blood of CRC patients and certain clinical features.

Predictors of performance: the relation of innovative dynamic capability and the services differentiation in retail

This scientific investigation studied the precedents of organizational performance, researching the relation between Innovative Capabilities and its relationship with Performance, and this relation with the mediation of Services Differentiation. For the qualitative methodology, aiming to identify the control items of the Service Differentiation measurement scale in the supermarket context. For the quantitative stage, the techniques of multivariate data analysis were used, such as Exploratory Factor Analysis, Confirmatory Factor Analysis, and Structural Equation Modeling, to assess the proposed model. The results found regarding the hypothesis tests demonstrated that the 1º hypothesis - Innovative capabilities have a positive relation with Services Differentiation was supported. 2º Hypothesis - Differentiation has a positive and significant relationship with performance, and it is therefore supported. 3º Hypothesis - Service Differentiation mediates the relationship between Innovative Capability and Performance - was supported.