No description available.

No description available.

published_or_final_version / abstract / Zoology / Doctoral / Doctor of Philosophy

Not available / text

Not available / text

Obesity is a common disease among all ages that has threatened human health and has become a global concern. Gut microbiota can affect human metabolism and thus may modulate obesity. Certain mixes of gut microbiota can protect the host to be healthy or predispose the host to obesity. Modern next-generation sequencing technique allows accessing huge amount of genetic information underlying microbiota and thus provides new insights into the functionality of these micro-organisms and their interactions with the host. Multiple previous studies have demonstrated that the microbiome might contribute to obesity by increasing dietary energy harvest, promoting fat deposition and triggering systemic inflammation. However, these researches are either based on lab cultivation studies or basic statistical analysis. In order to further explore how gut microbiota affect obesity, this thesis utilize a series of machine learning methods to analyze large amount of metagenomics data from human gut microbiome. The publicly available HMP (Human Microbiome Project) metagenomic sequencing data, contain microbiome data for healthy adults, including overweight and obese individuals, were used for this study. HMP gut data were organized based on two different feature definitions: taxonomic information and metabolic reconstruction information. Several widely used classification algorithms: namely Naive Bayes, Random Forest, SVM and elastic net logistic regression were applied to predict healthy or obese status of the subjects based on the cross-validation accuracy. Furthermore, the corresponding feature selection algorithms were used to identify signature features in each dataset that lead to the differences between healthy and obese samples. The results showed that these algorithms perform poorly on taxonomic data than metabolic pathway data though lots of selected taxa are still supported by literature. Among all the combinations between different algorithms and data, elastic net logistic regression has the best cross-validation performance and thus becomes the best model. In this model, several important features are found and some of these are consistent with the previous studies. Rerunning classifiers by using features selected by elastic net logistic regression again further improved the performance of the classifiers. On the other hand, this study uncovered some new features that haven't been supported by previous studies. The new features could also be the potential target to distinguish obese and healthy subjects. The present thesis work compares the strengths and weaknesses of different machine learning techniques with different types of features originating from the same metagenomics data. The features selected by these models could provide a deep understanding of the metabolic mechanisms of micro-organisms. It is therefore worth to comprehensively understand the differences of gut microbiota between healthy and obese subjects, and particularly how gut microbiome affects obesity.

Thesis (PhD (Physiological Sciences))--University of Stellenbosch, 2011. / Includes bibliography. / ENGLISH ABSTRACT: Genetic screening to improve obesity treatment outcomes is available despite the lack of conclusive evidence, specifically for Caucasian South Africans, in this regard. The aim of this study was to investigate the association between genotype (seven polymorphisms) and body mass index (BMI), health and lifestyle indicators in a cross-sectional sample of overweight/obese Caucasian adults (n=133), as well as the association between genotype and weight loss outcomes following an intervention (n=88) using a quasi experimental study design (time-series). The intervention consisted of a 24-week conservative weight loss programme that included dietary, physical activity and behavioural components. The primary null hypothesis for the cross-sectional sample, namely that there is no association between genotype and BMI, has not been rejected. A number of the secondary/exploratory hypotheses were rejected of which the most plausible associations (based on support by the literature and a physiological basis for the findng) are: 1) the mutant TT homozygotes of the GNB3 C825T polymorphism may have a higher risk to develop the metabolic syndrome (MetS) as they had significantly higher fasting triglyceride and glucose levels, a higher number of traits that met the diagnostic cut-off criteria for MetS and higher number of these subjects was diagnosed with MetS compared to the wild-type C-allele carriers; and 2) subjects with mutant alleles of either the FTO rs1421085 or rs17817449 polymorphisms may have poorer eating behaviours (a higher rigid control, habitual and emotional disinhibition, perceived hunger and internal locus for hunger) and higher intake of high-fat foods. The primary null hypothesis for the intervention sample, namely that there is no association between genotype and weight loss outcome, was not rejected for the FABP2 Ala54Thr, INSIG2 rs7566605, FTO rs1421085, ADRB3 Trp64Arg and GNB3 C825T polymorphisms. However, it was rejected in some instances indicating the following associations: 1) The wild-type TT homozygotes of the FTO rs17817449 polymorphism lost significantly more weight during the first two months of the program compared to the mutant allele carriers (this is a novel finding); 2) The wild-type Arg16Arg homozygotes of the ADRB2 Arg16Gly polymorphism lost significantly more weight during the first month of the program compared to the mutant allele carriers (this finding is supported by one other intervention study); 3) Subjects with a mutant C-allele of the INSIG2 rs7566605 polymorphism and a mutant Gly16-allele of the ADRB2 Arg16Gly polymorphism lost significantly less weight over the six month intervention period (this is a novel genegene interaction finding). A number of secondary/exploratory hypotheses were rejected, of which the most plausible finding include that the improvement in emotional disinhibition in the wild-type TT subjects of the FTO rs1421085 polymorphism was associated with a more pronounced decrease in BMI over the six month weight loss period. The integration of the results from this study with the literature indicates that there is insufficient evidence at this stage for genetic screening of the polymorphisms investigated in this study and the provision of evidence-based personalized recommendations for weight loss in obese individuals. It is recommended that these associations should be viewed as priority in future research. / AFRIKAANSE OPSOMMING: Genetiese sifting om die resultate van vetsug behandeling te verbeter is beskikbaar ten spyte van ‘n tekort aan genoegsame bewyse, spesifiek ten opsigte van Kaukasiërs van Suid-Afrika. Die doel van hierdie studie was om die assosiasie tussen genotipe (sewe polimorfismes) en liggaamsmassa indeks (LMI), gesondheid en lewenstyl indikatore in ‘n dwarssnit (cross-sectional) steekproef van oorgewig/vetsugtige Kaukasiër volwassenes (n=133) te ondersoek, asook die assosiasie tussen genotipe en gewigsverlies uitkomste na afloop van ‘n intervensie (n=88) in ‘n kwasi-eksperimentele studie ontwerp (tydreeks). Die intervensie het bestaan uit ‘n 24-week konserwatiewe gewigsverlies program met dieet, fisieke aktiwiteit en gedragskomponente. Die primêre nul hipotese vir die dwarsnit steekproef, naamlik dat daar geen assosiasie tussen genotipe en LMI is nie, is nie verwerp nie. ‘n Aantal sekondêre/spekulatiewe hipotesis is verwerp waarvan die mees geloofwaardige assosiasies (gebasseer op ondersteuning van die literatuur en ‘n fisiologiese basis vir die bevinding) die volgende insluit: 1) die mutante TT homosigote van die GNB3 C825T polimorfisme het moontlik ‘n hoër risiko vir die ontwikkeling van die metaboliese sindroom (MetS) aangesien hulle betekenisvolle hoër vastende trigliseriede en glukose vlakke gehad het, ‘n grooter aantal kenmerke gehad het wat aan die diagnostiese afsnykriteria vir MetS voldoen en ‘n grooter aantal van hierdie persone was met MetS gediagnoseer in vergelyking met die wilde-tipe C-alleel draers; en 2) persone met die mutante allele van die FTO rs1421085 of rs17817449 polimorfismes het moontlik ‘n swakker eetgedrag (‘n hoër rigiede kontrole, gewoonte en emosionele disinhibisie, waarneembare honger en interne lokus van honger) en ‘n hoër inname van hoë-vet voedsel. Die primêre nul hipotese vir die intervensie steekproef, naamlik dat daar geen assosiasie tussen genotipe en gewigsverlies uitkomste is nie, is nie vir die FABP2 Ala54Thr, INSIG2 rs7566605, FTO rs1421085, ADRB3 Trp64Arg en GNB3 C825T polimorfismes verwerp nie. Dit was egter in sommige gevalle vir die volgende assosiasies verwerp: 1) Die wilde-tipe TT homosigote van die FTO rs17817449 polimorfisme het betekenisvol meer gewig in die eerste twee maande van die program verloor in vergelyking met die mutante alleel draers (dit is ‘n nuwe bevinding); 2) Die wilde-tipe Arg16Arg homosigote van die ADRB2 Arg16Gly polimorfisme het betekenisvol meer gewig gedurende die eerste maand van die program verloor in vergelyking met die mutante alleel draers (hierdie bevinding word ondersteun deur een ander intervensie studie); 3) Persone met ‘n mutante C-alleel van die INSIG2 rs7566605 polimorfisme en ‘n mutante Gly16-allele van die ADRB2 Arg16Gly polimorfisme het minder gewig tydens die ses maande intervensie periode verloor (dit is ‘n nuwe geen-geen interaksie bevinding). ‘n Aantal sekondêre/ spekulatiewe hipoteses is verwerp, waarvan die mees geloofwaardigste bevinding insluit dat ‘n verbetering in emosionele disinhibisie van die wild-tipe TT persone van die FTO rs1421085 polimorfisme geassosieer was met ‘n meer prominente daling in LMI oor die ses maande gewigsverlies periode. Die integrasie van die resultate van hierdie navorsing met die literatuur dui aan dat daar op hierdie stadium onvoldoende bewyse vir genetiese sifting en die voorsiening van bewys-gebasseerde persoonlike aanbevelings vir gewigsverlies in vetsugtig individue bestaan vir die polimorfismes wat ondersoek is. Dit word aanbeveel dat hierdie assosiasies as prioriteit in toekomstige navorsing beskou moet word.

Thesis (PhD)--Stellenbosch University, 2012. / Includes bibliography / ENGLISH ABSTRACT: Background/Aims: Obesity has increased rapidly in South African children and adolescents with significant variability observed among racial groups. Genes that regulate appetite have been studied in different populations worldwide, but their role in obesity among South African adolescents is unknown. The present study aimed at investigating the role of these genes, and their combined effect with physical activity in the development of obesity among South African adolescents. Methods: A total of 1564 South African school learners of Caucasian (n= 146), Mixed Ancestry (n= 872) and Black African (n= 537) ethnic groups were recruited for a research project that aimed to elucidate diabetes and the metabolic syndrome in children and adolescents attending schools in periurban areas of the Western Cape. The present case-control study included 227 obese-overweight (115 Black Africans and 112 Mixed Ancestry), and 204 normal weight (94 Black Africans and 110 Mixed Ancestry) adolescents learners. The learners were genotyped for nine polymorphisms (LEP: 19G>A, Lys36Arg, Val94Met; LEPR: Lys109Arg; Gln223Arg, Lys656Asn; CART: c.160-33G>A, c.499delA, and c.517A>G; GHRL: Leu72Met; and MC3R: Thr6Lys, Val81Ile) using allele-specific restriction enzyme analysis and automated sequencing. Genotype and haplotype associations with anthropometric variables such as body mass index (BMI), waist, hip, and mid-upper-arm circumferences (WC, HC, MUAC), and metabolic traits (fasting blood glucose, high density lipoproteincholesterol, total cholesterol), and blood pressure were further conducted. Furthermore, the type and frequency of physical activity was assessed by means of structured questionnaires; and its effect on obesity-related variables investigated in learners that were genotyped for the MC3R Thr6Lys and Val81Ile polymorphisms. Results: In a stepwise backward logistic regression analysis (containing age, gender, and LEP, LEPR, CART and GHRL polymorphisms), CART c.517A>G was independently significantly associated with obesity (OR= 5.98; 95%CI= 2.02, 21.27). CART c.517G carriers had higher MUAC (b coefficient= 1.88; 95%CI= 0.31, 3.44) while the LEPR 109Arg allele was significantly associated with decreased BMI (b coefficient = -2.36; 95%CI= -4.24, -0.47), WC (b coefficient = -5.66; 95%CI= -9.89, -1.44) and MUAC (b coefficient = -1.61; 95%CI= -3.00, -0.22); after adjusting for age, gender, and ethnicity. The haplotype containing the three LEP polymorphisms (A-A-A compared to the reference G-A-G haplotype) increased BMI (p= 0.0155), MUAC (p= 0.0146), and HC (p= 0.0128). The minor alleles of the MC3R polymorphisms decreased BMI, HC, WC, MUAC and TC; whilst only the Thr6Lys was associated with systolic and diastolic blood pressure (p= 0.0047 and 0.0027, respectively) in Mixed Ancestry learners. Doing house chores was associated with lower total cholesterol, independently and in the presence of the 81Ile allele (b coefficient = -0.355; 95%CI= 0.148, 0.561). Conclusion: To our knowledge, this is the first study that reports CART c.517A>G polymorphism as a risk factor for obesity in adolescents. Furthermore, the present study demonstrated that the MC3R polymorphisms had a positive effect on total cholesterol, which was further enhanced in physically active individuals. Similar to other studies, LEPR Lys109Arg and LEP polymorphisms were associated with variations in obesity-related variables among Black African and Mixed Ancestry South African learners. / AFRIKAANSE OPSOMMING: Agtergrond/Doelwitte: Vetsug het drasties toegeneem in Suid-Afrikaanse kinders en adelossente met ‘n beduidende variasie opgemerk tussen verskillende rassegroepe. Gene verantwoordelik vir regulering van eetlus is reeds wêreldwyd in verskillende bevolkingsgroepe bestudeer, maar hul rol in oorgewig Suid-Afrikaanse adolessente is onbekend. Die huidige studie was daarop gerig om ondersoek in te stel na die rol van hierdie gene en hul gekombineerde effek met fisiese aktiwiteit in die ontwikkeling van vetsug onder Suid-Afrikaanse adolessente. Metodes: ‘n Totaal van 1564 Suid-Afrikaanse leerders van Kaukasiese Afkoms (n=146), Gemengde Afkoms (n=872) en Swart Afkoms (n= 537) was gewerf in die navorsingsprojek wat ten doel gehad het om kinders en adolosente met diabetes en die metaboliese sindroom te identifiseer wat skole bygewoon het in semi-voorstedelike gebiede van die Wes-Kaap. Die huidige gevalle studie het 227 vetsugtige-oorgewig (115 Swart Afkoms en 110 Gemengde Afkoms) en 204 normale gewig (94 Swart Afkoms en 110 Gemengde Afkoms) leerders ingesluit. Die leerders was gegenotipeer vir nege polimorfismes (LEP: 19G>A, Lys36Arg, Val94Met; LEPR: Lys109Arg; Gln223Arg, Lys656Asn; CART: c.160-33G>A, c.499delA, and c.517A>G; GHRL: Leu72Met; and MC3R: Thr6Lys, Val81Ile) met die gebruik van alleel-spesifieke restriksie ensiem analises en geoutomatiseerde DNA volgorde bepalings tegnieke. Genotipiese en haplotipiese assosiasies met antropometriese veranderlikes soos liggaamsmassa indeks (BMI), middel-, heup- en mid-boarm omtrek (WC, HC, MUAC), metaboliese tendense (vastende bloed glukose, hoë-digtheid lipoproteïen-cholesterol, totale cholesterol) en bloeddruk was ook uitgevoer. Die tipe en frekwensie fisiese aktiwiteit was geassesseer deur middel van gestruktureerde vraelyste; en die uitwerking daarvan op vetsugverwante veranderlikes ondersoek in leerders wat vir die MC3R Thr6Lys en Val81Ile polimorfismes gegenotipeer was. Resultate: Statistiese ontleding (‘‘stepwise backward logistic regression analysis”), wat ouderdom, geslag en polimorfismes (LEP, LEPR, CART GHRL) ingesluit het, het getoon dat CART c.517A>G betekenisvol onafhanklik geassosiasieer was met vetsug (OR= 5.98; 95% CI= 2.02, 21.27). CART c.517G draers het ‘n hoër MUAC waarde gehad (b koeffisient = 1.88; 95%CI= 0.31, 3.44), terwyl die LEPR 109Arg alleel betekenisvol geassosieer was met verlaagde BMI ((b koeffisient = -2.36; 95%CI= -4.24, -0.47), WC (b koeffisient = -5.66; 95%CI= -9.89, -1.44) en MUAC (b koeffisient = -1.61; 95%CI= -3.00, -0.22) na die aanpassing van ouderdom, geslag en etnisiteit. Die haplotipe met die drie LEP polimorfismes (A-A-A teenoor die G-A-G verwysingshaplotipe) het die BMI (p= 0.0155), MUAC (p= 0.0146) en HC (p= 0.0128) verhoog. Die mindere allele van die MC3R polimorfismes het die BMI, HC, WC, MUAC en TC verlaag; terwyl slegs die Thr6Lys polymorfisme met sistolies en diastolies bloeddruk (p= 0.0047 en p= 0.0027, onderskeidelik) geassosieer was in Gemengde Afkoms leerders. Die verrigting van algemene huistake was geassosieer met laer totale kolesterol vlakke, onafhanklik en in die teenwoordigheid van die 81lle alleel (b koeffisient= -0.355; 95%CI= 0.148, 0.561). Gevolgtrekking: Na ons wete is hierdie die eerste studie wat die CART c.517A>G polimorfisme as ‘n risikofaktor vir vetsug in adolessente aantoon. Die huidige studie toon ook dat die MC3R polimorfisme ‘n positiewe effek op totale kolesterol gehad het, wat ook verder versterk was in fisiese aktiewe individue. Soortgelyk aan ander studies, was die LEPR Lys109Arg en LEP polimorfismes geassosieer met variasies in vetsug-verwante veranderlikes onder Suid-Afrikaanse Swart en Gemengde Afkoms leerders. / This research was supported by a grant from the University Research Fund of the Cape Peninsula University of Technology, Harry Crossley, University of Stellenbosch, Faculty of Health Sciences, Medical Research Council, and the National Health Laboratory Services, South Africa.