Filipe Macedo Gudin

Ensaio Opcional (10/03)

The discovery of structures and molecules involved in inheritance of features shed light into genetics and many other science fields. However, understanding how these molecules are connected in the process of inheritance is the great contribution of the central dogma of molecular biology. The idea of this dogma is mainly flow of information, i.e, how the information can be properly transfered from one source to another. Basically, DNA is the main source of information, which is copied into RNA and, finally to a particular protein. There are many other genetic, and even epigenetic, pathways to transfer information, though. Although this diversity of mechanisms is high, the universality of this dogma is beyond our understanding. First, the comparison of genetic material from many different species was made viable, because the genetic code of nucleotides are the same. The process of flow of information are also present both in local (e.g., cell division) and historical levels (e.g., evolution of phenotypes). Furthermore, even the flow of information from nucleotides to amino acids are global (despite a few particularities in some taxa). Therefore, there is no doubt that the formulation of the central dogma of molecular biology revolutionized the way we think about molecular process and evolution.

Acredito que a sentença- chave seria começar com: "the great contribution of the central dogma of molecular biology is the understanding…"
De resto achei que estava bem estruturado (sentenças claras objetivas, com seus devidos pontos finais)
Corrigido por Maila Beyer

Ensaio 2 (24/03)

Genetic drift and natural selection are the two main process involved in fixation of mutations in genetic material. Nonetheless, the impact of the first had been historically underestimated until the work of Kimura pointed it out. Many evolutionists and geneticists had treated natural selection as the leading process introducing variation in populations. This kind of thinking inevitably leads to the concept of adaptation, i. e., all variation is important for the fitness of the organism. However, Kimura made a very important contribution when analyzing available evidences of variation in nucleotides and proteins. Based on these observations and some estimates of general rates of substitution, he argues that natural selection alone could not produce such huge diversity. Genetic drift is, thus, brought back as a better explanation, suggesting that the majority of this mutations are neutral or nearly neutral. Some examples offered to clarify this issue are the degenerated code, especially the third position of the codon, and the substitution of amino acids in hemoglobins that do not alter its main function. Genetic drift, therefore, is not just some process, but it seems to be the principal cause of variation in populations, even more than natural selection. Additionally, the consequences derived from this explanation are also huge. Most of genetic (and even phenotypic!) variation is not an adaptation. They can just be irrelevant to the organism.

Comentário por Pietro Vicari
A linguagem do texto está clara e bastante objetiva. As sequências me parecem claramente conectadas o que o torna o texto fluido.
Senti falta de citações que suportassem as afirmações contidas no texto. Na esfera conceitual um destaque para a frase "Many evolutionists and geneticists had treated natural selection as the leading process introducing variation in populations".
A seleção não introduz as variações nas populações, apenas seleciona a variabilidade que surge por outros processos. No mais considero seu ensaio bom e acredito que aprendi fazendo a revisão.

Ensaio 3 (31/03)

The role and intensity of both genetic drift and selection in molecular evolution have been intensively debated in the last decades. However, is it possible to verify which is more frequent in a given system? The effective population size (Ne) is an important concept used to understand the effects of these two process. Basically, Ne can be defined as the individuals of a given population that contribute to the genetic composition of the next generation. There are many factors that influence Ne, such as sex ratio, inbreeding, frequency and inheritance of alleles, changes in population size, and geographic or genetic subpopulations. There are many models that incorporate these factors to calculate Ne. Nevertheless, the generic effect of Ne in genetic drift and selection can be easily understood. Populations with low Ne are submitted more intensively to genetic drift. Although genetic drift still acts in populations with higher Ne values, selection tends to be more effective in fixation of alleles.

Comentários:
De modo geral o texto está bom e bem claro, com um bom uso dos conectivos.
Você começa duas frases seguidas com a mesma expressão "There are many…". Acho que poderia juntar as duas frases, tipo "…genetic subpopulations, as well as many models that…". Na minha opinião ficaria mais fluido e menos trucada essa parte.

Ensaio 4 (05/05)

Tree thinking brought a new perspective to evolutionary studies, due mainly to two concepts: common ancestry and descent. All life forms are derived from a common ancestor, sharing characters derived from descent. Additionally, the development of phylogenetic systematics enabled a new way to incorporate these concepts in evolutionary studies and to fix tree thinking in many approaches in Biology (and other fields where these factors of heredity apply to). Phylogenetic trees can be used as background to understand evolutionary processes, classify organisms and many other questions in science. However, clear understanding of what a tree represents seems to be lacking for many users of phylogenetic trees. First, trees are hypothesis of relationships between taxa, exhibiting how close or not a taxon is related to another based on descent and common ancestry. Since trees are hypothesis, they can not be treated as true representations of an evolutionary pattern. Furthermore, trees are inferred from characters, which are prone to bias, subjectivity and constraints in the process of data gathering. Different treatments or interpretations of characters can lead to different trees, i.e., different hypothesis. Both these issues must be considered when one wants to work with phylogenetic trees.

Bom texto, gostei da sentencia tópico e do jeito em que você ordeno, ligou e transmitiu suas ideias. Talvez, dependendo de seu publico alvo, seria necessário definir em poucas palavras o conceito de sistemática filogenética. Eu reduziria a oração “Additionally, the development…”, terminando ela em: "…in many approaches in biology and other fields of science."

Ensaio 5 (12/05)

Distance algorithms are commonly used nowadays to explore molecular data in comparative studies However, these phenetic approaches were, primarily, proposed for phylogenetic reconstructions, i.e., inferences about taxa relationships. The number of phylogenetic trees exponentially increases with the quantity of organisms included in an analysis, generating astronomic numbers of possible trees (just like big forests). These algorithms enable fast and optmized explorations of possibilities inside these forests. Nevertheless, use of distance algorithms for phylogenetic reconstructions were severely criticized. They are based on total similarity (or more specifically on dissimilarity) of organisms, ignoring transformations occurred on each character. Additionally, as the construction of the matrix (and the tree) is made pairwise, the final topology is extremely dependent on the first taxon selected. Despite distance algorithms are not recommended for phylogenetic reconstructions they are still useful nowadays, especially with the increasing amount of genomic data.

Comentário por Pietro Vicari
Olá Filipe. Parabéns pelo texto. Acredito que os princípios de escrita científicas indicados foram seguidos. Existem apenas alguns problemas com pontuações, principalmente no sentido de quebrar muitos as frases. Por exemplo a frase "Additionally, as the construction of the matrix (and the tree) is made pairwise, the final topology is (…)". A frase é quebrada para incluir a informação "as the construction of the matrix is made pairwise", e essa informação é quebrada para incluir a informação "(and the tree)". Poderia tentar mudar a ordens que expõe algumas coisas para que fique mais direto. Além disso achei um pouco confuso sobre sua opinião quanto ao uso desses métodos. Você diz que não são recomendados mas são úteis especialmente com o crescimento de dados genômicos. Essa relação poderia ser explica um pouco melhor para mostrar claramente o que você quer dizer.

Ensaio 6 (19/05)

A análise de dados moleculares em sistemática filogenética foi fortemente impactada com a abordagem de verossimilhança, tanto em complexidade quanto em custo de processamento de dados. No contexto de sistemática filogenética, de forma geral, verossimilhança estima a melhor explicação (topologia) que poderia ter gerado os dados coletados (matriz de sequências alinhadas), baseada em um modelo de substituição molecular. Os valores dos parâmetros presentes no modelo utilizado são estimados frente a uma topologia onde os caracteres são otimizados. No final, a estimativa que apresentar a maior verossimilhança será escolhida como a solução ideal para a sua base de dados. Dessa forma, a abordagem de verossimilhança permite implementar em uma inferência filogenética, por meio dos modelos de substituição, diversos conceitos sobre evolução molecular, levando a uma forma de lidar com dados moleculares mais coerente com as teorias vigentes.

Ensaio 7 (26/05)

Despite all models are simplistic representations of reality, some are more useful than others. This principle is also present in phylogenetic inferences based on statistical criteria. To search for a topology that better optimize a character matrix of molecular sequences, a model of substitution is necessary. However, the number of parameters present in a model need to be evaluated and estimated, specially when choosing the most useful model for an analysis. There are many types of parameters in models of molecular substitution. The first type of parameter is the rate of substitution between nucleotides. This rate may vary between being the same for all nucleotides (i.e., just one parameter) to even different for all of them (i.e., twelve parameters, which is the maximum possibility considering four nucleotides). Additionally, there are more parameters that can alter and correct this rate. Frequency of nucleotides (F), variance in substitution rates between sites (gamma distribution) and number of invariable sites (I) in a sequence can dramatically alter the value of rates, and need to be included for correction. All these parameters need to be estimated when choosing the most useful model before searching for the best tree in a phylogenetic analysis.

Correção Maila Beyer
Boa frase inicial, trazendo a idéia geral do texto.
frases claras e no geral curtas.
o texto traz um fechamento.
faltou referenciar as idéias que trouxe no texto.

Ensaio 8 (01/06)

In phylogenetic systematics the total amount of possible trees grows exponentially with the number of taxa included in an analysis. This huge tree space of possibilities prevents an examination of all trees to find the best. However, there are many algorithms to heuristically explore this tree space. A tree space is complex and its configuration can present one or many local optima, i.e., hills of your tree space containing best trees. These hills can present different heights, though, which means that the best trees will be at the top of the highest hill. Search algorithms can ensure that you can find a local optima in your tree space and, consequently, the best tree. This is known as hill-climbing. Nonetheless, the first search algorithms proposed (NNI, SPR and TBR) may get stuck in some local optima that do not contain the best tree in your tree space (remember the difference of hills' heights mentioned above). So, there are more sophisticated algorithms that can change the surface of tree space or force your search to jump from one hill to another. The efficacy of these algorithms combined is what enabled the development of phylogenetic systematics in the last decades.

comentado por Marcos

Olá Filipe.

Muito bom seu ensaio. Não tenho nada para apontar.
Eu só colocaria como primeira frase (topic sentense) algo que evidencia-se mais o uso de algoritmos como ferramenta de busca de árvores, mas não tem nenhum problema com a sua frase.

abraço

Ensaio 9 (09/06)

Molecular clock is an useful and controversial approach for evolutionary studies. Through estimates of molecular clock models, relative dates of divergence between lineages can be obtained. Additionally, these dates can be posteriorly calibrated (e.g., with fossil data or age of terminals), generating absolute dates of divergence. However, there are many issues that need to be accounted for by including molecular clock models and calibration in a phylogeny. Depending on the clock model, estimated ages can be wrong and sometimes even nonsensical, since the genes used do not fit that model (e.g., strict clock model in genes with varied rates of substitution). Furthermore, the point of calibration can change the age of all nodes of a tree, which demands extra careful where to insert it. Although the estimation of these dates is both debatable and extremely sensitive to variation of parameters and data, dates of divergence can be extremely elucidative in historic biogeography, reconstruction of ancestral character states, and other evolutionary areas in biology.

Ensaio Final e Avaliação da Disciplina

Entrei na disciplina para poder compreender melhor a área de dados moleculares em sistemática filogenética, mas me surpreendi ao perceber que o campo de evolução molecular é muito mais abrangente que a minha área de atuação. Achei a estrutura da disciplina muito boa e inovadora. O fato de desenvolver a escrita em pequenos ensaios ajuda a fixar o que foi aprendido, além de desenvolver a habilidade de escrever sobre algo em qualquer momento. A forma de avaliação eu também achei muito boa, inclusive o sorteio dos pontos, pois me levou a estudar bem mais outros assuntos que, normalmente, não me esforçaria para me aprofundar. Além disso, a leitura de artigos foi muito boa e interessante. Acredito que saí com uma bagagem de conteúdo e compreensão a respeito do campo e contexto científico em que estamos (tipo de escrita, impacto de diferentes áreas da ciência na comunidade científica, literatura de outras áreas) bem maior do que eu tinha. Certamente pude suprir algumas falhas na minha formação básica durante essa disciplina.

Com relação a sugestões, acredito que seria ideal liberar os artigos com mais antecedência na semana, como na segunda ou terça-feira, por exemplo. Ao mesmo tempo, compreendo que a disciplina foi ministrada pela primeira vez e o desenvolvimento do conteúdo e escolha dos artigos foi ocorrendo ao longo do semestre. Em uma segunda edição, acredito que já será possível disponibilizar o material mais cedo para os alunos terem tempo de organizar suas leituras. Acho importante também incluir uma aula específica sobre inferência bayesiana, pois nenhuma disciplina do departamento aborda esse assunto. Como é um método complexo e extremamente presente na literatura, seria interessante ter uma aula para explicar os conceitos básicos, tanto para crítica quanto para o uso do método.

Para a autoavaliação, considero que minha nota seja 1,0.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License