The work undertaken within this thesis is towards the development of a representative set of sequence driven features for the prediction of structural classes of proteins. Proteins are biological molecules that make living things function, to determine the function of a protein the structure must be known because the structure dictates its physical capabilities. A protein is generally classified into one of the four main structural classes, namely all-α, all-β, α + β or α / β, which are based on the arrangements and gross content of the secondary structure elements. Current methods manually assign the structural classes to the protein by manual inspection, which is a slow process. In order to address the problem, this thesis is concerned with the development of automated prediction of structural classes of proteins and extraction of a small but robust set of sequence driven features by using the amino acid indices. The first main study undertook a comprehensive analysis of the largest collection of sequence driven features, which includes an existing set of 1479 descriptor values grouped by ten different feature groups. The results show that composition based feature groups are the most representative towards the four main structural classes, achieving a predictive accuracy of 63.87%. This finding led to the second main study, development of the generalised amino acid composition method (GAAC), where amino acid index values are used to weigh corresponding amino acids. GAAC method results in a higher accuracy of 68.02%. The third study was to refine the amino acid indices database, which resulted in the highest accuracy of 75.52%. The main contributions from this thesis are the development of four computationally extracted sequence driven feature-sets based on the underused amino acid indices. Two of these methods, GAAC and the hybrid method have shown improvement over the usage of traditional sequence driven features in the context of smaller and refined feature sizes and classification accuracy. The development of six non-redundant novel sets of the amino acid indices dataset, of which each are more representative than the original database. Finally, the construction of two large 25% and 40% homology datasets consisting over 5000 and 7000 protein samples, respectively. A public webserver has been developed located at http://www.generalised-protein-sequence-features.com, which allows biologists and bioinformaticians to extract GAAC sequence driven features from any inputted protein sequence.

Na Inferência Estatística, é comum, após a realização de um experimento, testar simultaneamente um conjunto de diferentes hipóteses de interesse acerca de um parâmetro desconhecido. Assim, para cada hipótese, realiza-se um teste de hipótese e, a partir disto, conclui-se algo sobre os parâmetros de interesse. O objetivo deste trabalho é avaliar a (falta de) concordância lógica entre as conclusões obtidas a partir dos testes realizados após a observação de um único experimento. Neste estudo, é apresentada uma definição de classe de testes de hipóteses, uma função que para cada hipótese de interesse associa uma função de teste. São então avaliadas algumas propriedades que refletem como gostaríamos que testes para diferentes hipóteses se comportassem em termos de coerência lógica. Tais propriedades são exemplificadas através de classes de testes que as satisfazem. A seguir, consideram-se conjuntos de axiomas para classes. Estes axiomas são baseados nas propriedades mencionadas. Classes de testes usuais são investigadas com relação aos conjuntos de axiomas propostos. São também estudadas propriedades advindas de tais conjuntos de axiomas. Por fim, estuda-se um resultado que estabelece uma espécie de conexão entre testes de hipóteses e estimação pontual. / In Statistical Inference, it is usual, after an experiment is performed, to test simultaneously a set of hypotheses of interest concerning an unknown parameter. Therefore, to each hypothesis, a statistical test is performed and a conclusion about the parameter is drawn based on it. The objective of this work is to evaluate the (lack of) logical coherence among conclusions obtained from tests conducted after the observation of a single experiment. In this study, a definition of class of hypotheses tests, a function that associates a test function to each hypothesis of interest, is presented. Some properties that reflect what one could expect (in terms of logical coherence) from tests to different hypotheses are then evaluated. These properties are exemplified by classes of hypotheses tests that respect them. Then, sets of axioms based on the properties studied are proposed to classes of hypotheses tests. Usual classes of hypotheses tests are investigated with respect to these sets of axioms. Some properties related to these sets of axioms are then analyzed. At last, a result which seems to connect hypotheses testing and point estimation is stated.

Na Inferência Estatística, é comum, após a realização de um experimento, testar simultaneamente um conjunto de diferentes hipóteses de interesse acerca de um parâmetro desconhecido. Assim, para cada hipótese, realiza-se um teste de hipótese e, a partir disto, conclui-se algo sobre os parâmetros de interesse. O objetivo deste trabalho é avaliar a (falta de) concordância lógica entre as conclusões obtidas a partir dos testes realizados após a observação de um único experimento. Neste estudo, é apresentada uma definição de classe de testes de hipóteses, uma função que para cada hipótese de interesse associa uma função de teste. São então avaliadas algumas propriedades que refletem como gostaríamos que testes para diferentes hipóteses se comportassem em termos de coerência lógica. Tais propriedades são exemplificadas através de classes de testes que as satisfazem. A seguir, consideram-se conjuntos de axiomas para classes. Estes axiomas são baseados nas propriedades mencionadas. Classes de testes usuais são investigadas com relação aos conjuntos de axiomas propostos. São também estudadas propriedades advindas de tais conjuntos de axiomas. Por fim, estuda-se um resultado que estabelece uma espécie de conexão entre testes de hipóteses e estimação pontual. / In Statistical Inference, it is usual, after an experiment is performed, to test simultaneously a set of hypotheses of interest concerning an unknown parameter. Therefore, to each hypothesis, a statistical test is performed and a conclusion about the parameter is drawn based on it. The objective of this work is to evaluate the (lack of) logical coherence among conclusions obtained from tests conducted after the observation of a single experiment. In this study, a definition of class of hypotheses tests, a function that associates a test function to each hypothesis of interest, is presented. Some properties that reflect what one could expect (in terms of logical coherence) from tests to different hypotheses are then evaluated. These properties are exemplified by classes of hypotheses tests that respect them. Then, sets of axioms based on the properties studied are proposed to classes of hypotheses tests. Usual classes of hypotheses tests are investigated with respect to these sets of axioms. Some properties related to these sets of axioms are then analyzed. At last, a result which seems to connect hypotheses testing and point estimation is stated.

Today's automotive requirements have resulted in complex Sheet Metal Forming (SMF) processes of Sheet Metal (SM) with reduced formability, and thus it is crucial to be able to predict formability accurately to prevent material failure during SMF. Formability predictions today utilize Forming Limit Curves (FLC)s in Finite Element Analysis (FEA), but  FLCs are not valid for the Non-Linear Strain Paths (NLSP)s generated during SMF. One purpose of this thesis is thus to increase the knowledge on FP handling NLSP,  which was obtained through providing suggestions of failure models for handling NLSP effects, based upon literature on the subject. Generating NLSP experimentally is both time and material costly with the conventional method, thus the second purpose of this thesis was to increase the knowledge on test procedures for generating NLSP in SM. Based upon the findings of Chandramohan \cite{chandramohan_study_2021} five test procedures for generating NLSP were put forward, and the Nakajima test with modified punch geometry was chosen for further study.   In this thesis, the NLSP characteristics of two modified punch geometries were evaluated by FEA performed using LS-DYNA. For the FEA three specimens with blank width of 50, 100 and 200 mm was used, and the anisotropic Barlat yld2000  was used as the material model. This material model was calibrated to material data of Mild steel CR4, Aluminium alloy AA6016, and Dual-phase steel DP800. The results for all materials showcased similar reacquiring general NLSP characteristics at the corners of the punch features, which are unfavorable positions when failure by necking is evaluated, and thus it was concluded that the tested punch geometries are not favorable and more development of the punch geometry is needed. / Dagens fordonskrav, har lett till komplexa plåtformnings processer av plåtmaterial med reducerad formbarhet, och det är därför väsenligt att kunna förutsäga formbarhet noggrant för att förhindra materialbrott under plåtformning. Försträckning och brott förutses idag genom Formgränskurvor (FGK) i finita element analyser (FEA), men dessa gäller inte för icke-linjära töjningsvägar som uppkommer under plåtformning. Ett syfte av denna avhandling är därför att öka kunskapen kring modeller för att förutsäga formbarhet under icke-linjära töjningsbanors effekter, vilket uppnådes genom att  presenteras  förslag på brott modeller för att hantera de icke-linjära töjningsvägar baserade på  literatur inom området. Att generera icke-linjära töjningsvägar experimentellt är både tids och materialkrävande med den konventionella metoden, således är det andra syftet av denna avhandling att öka kunskapen kring test metoder för att generera icke-linjär töjningsbvägar i plåt. Baserat på Chandramohans \cite{chandramohan_study_2021} resultat diskuteras fem test procedurer för att generera icke-linjära töjningsvägar, och Nakajima test med modifierad stämpelgeometri valdes för vidare studie.  I denna avhandling studerades töjningsignaturen av två stämpelgeometrier med FEA i LS-DYNA. Till FEA:n användes tre ämnen med bredd av 50, 100 och 200mm, och anisotropiska Barlat yld2000 användes som materialmodell. Denna materialmodell kalibrerades mot experimentella mätvärden för mjukt stål CR4, Aluminiumlegering AA6016 och Stål DP800. Resultaten visade för alla material återkommande generella icke-linjära töjningsbanor enbart för hörnorna på stansgeometrierna, vilket är icke önskvärda positioner då brott pga. midjebildning utvärderas, och således drogs slutsatsen att nuvarande stansgeometri inte är gynnsam och ytterligare utveckling behövs.