Caio Gueratto Coelho Da Silva

Ensaio 1 opcional (10/03/2017)

Central dogma of molecular biology and its implications to modern phylogenetic inference.

During the class, we discussed some basic aspects of the so called Central Dogma of molecular biology, postulated by Crick in 1958, and becoming a well-established point in the Biology. This dogma had new information added, but still with minor changes. Beyond that, we discussed the implications of the dogma to the phylogenetic inference. After this discussion, the class treated about the genetic code and its properties (e.g. formed by 4 nucleotides, which are organized in groups of three (codons) and these groupings can form the 20 aminoacids) as well as the their implication to evolutionary reconstructions. The point that most called my attention was the fact that with the rising of the molecular biology and its central dogma, it became possible to compare groups of individuals that are not comparable using morphology and anatomy. The comparison is impractical due to the huge evolutionary time that separates these groups, what makes impossible to see morphological homologies between them. One example is the construction of phylogenetic relashionship of all the living eukaryotes, as the one presented in Baldauf (2008). With this new approach of comparing nucleotides instead of morphological characters, the relationship between all living individuals is beginning to be elucidated. And this thanks to the central dogma of molecular biology.

Comentarios

Desculpas com antecipação pelo meu português

Achei interesante a ideia de fazer uma revisão da aula mesma desde sua perspectiva. E importante que, se faz essa escolha, a voz seja constante ao longo do paragrafo (Ao final muda de primeira a terceira voz). Incluo comentarios particulares a continuação:

During the class, we discussed some basic aspects of the so called Central Dogma of molecular biology, postulated by Crick in 1958, and becoming a well-established point in the Biology. This dogma had new information added, but still with minor changes.

Esta sentença trata dois negócios diferentes: o topico da aula e o estabelecimento do dogma central. Esto poderia ser dividido em duas o tres sentenças. E importante manter a constante de uma ideia por sentença.

This dogma had new information added, but still with minor changes

nova informação adicionada a o que?

Beyond that, we discussed the implications of the dogma to the phylogenetic inference. After this discussion, the class treated about the genetic code and its properties (e.g. formed by 4 nucleotides, which are organized in groups of three (codons) and these groupings can form the 20 aminoacids) as well as the their implication to evolutionary reconstructions.

A secção em suportes tira a atenção tira a atenção do fluxo da sentença. Ficaría melhor como uma sentença aparte.

The point that most called my attention was the fact that with the rising of the molecular biology and its central dogma, it became possible to compare groups of individuals that are not comparable using morphology and anatomy

Tem algums erros idiomaticos aquí, uma sugestão:

''The point that called my attention the most was that, with the rise of molecular biology, came the possibility of comparing groups which were not comparable using morphology or anatomy.''

The comparison is impractical due to the huge evolutionary time that separates these groups, what makes impossible to see morphological homologies between them. One example is the construction of phylogenetic relashionship of all the living eukaryotes, as the one presented in Baldauf (2008). With this new approach of comparing nucleotides instead of morphological characters, the relationship between all living individuals is beginning to be elucidated.

Pode simplificar

And this thanks to the central dogma of molecular biology.

Esta sentença é redundante.

Bom trabalho!

(Comentarios por Juan Jiménez)

Ensaio 2 (17/03/2017)

Rapidly pulishing molecular data versus traditional morphological taxonomy: a heated debate.

Species delimitation is a very debated biological topic, specially on the matter of candidate species. The income of molecular-based techniques opened way to publish data regarding candidate species (species that possibly are undescribed and, therefore, unknown until the moment the data were published) very rapidly. On the other hand, it also gave rise to critiques from taxonomists. They support that this new approach leave a lot of pending questions regarding whether this candidate species are in fact new species or not. Beyond that, taxonomists claim that more work should be done before publishing a list of possible new species to be sure that they are in fact new ones. However, the supporters of the new approach claim that the readiness of published data is important to have at least something at hand about a taxon, once this can be used to create conservational measures to protect the habitat in which the taxon lives in. Therefore, in the last instance, the species delimitation work would help to protect natural areas from being destroyed. After accomplishing this, one could easily do taxonomical work anytime, once the taxon and its habitat would be preserved. My personal opinion is that biology researchers have this ethical demand of trying to prevent as much environment destruction as we can. If this quick works help us to do prevent areas from destruction, then we should at least try and afterwards do more extensive work on the area (or taxon).

Comentários (Darlan Redü)

A sentença tópico deveria ser outra. Seu título remete a um debate muito mais amplo do que a delimitação de espécies.

Cuidado com expressões repetidas (candidate species na primeira e segunda frase).

However, the supporters of the new approach claim that the readiness of published data is important to have at least something at hand about a taxon, once this can be used to create conservational measures to protect the habitat in which the taxon lives in.
Essa sentença poderia ser dividida em duas: "…about a taxon. This can be…"

Therefore, in the last instance, the species delimitation work would help to protect natural areas from being destroyed.
Eu tiraria o "in the last instance" para deixar a frase mais direta. Bom uso do conectivo. Será que essa frase e a anterior não são repetitivas? Para mim elas dizem a mesma coisa.

Acho que a ultima frase é redundante também. Você já falou sobre preservação de habitats e manutenção do material para estudos posteriores. Se retirar essa frase, seu ensaio terminaria com sua opinião pessoal, que particularmente acho um bom fechamento para o ensaio.

O ponto de vista do big data foi explicado e bem defendido pela sua opinião pessoal. Porém, acho que a crítica dos taxonomistas não. O ponto principal da crítica não é apenas se as espécies são realmente novas ou não, mas sim que essas práticas não acompanham a filosofia e os métodos da sistemática tradicional.

Ensaio 3 (24/03/2017)

Stochastic changes as the main mutational genomic process.

The knowledge we have nowadays regarding mutational process in genomes is that they are mainly caused by genetic drift. This was proposed by Kimura (1968), when he studied nucleotide and amino acid substitutions in hemoglobin molecules. Before his work, people believed that only a few changes occurred in the hemoglobin coding genes throughout the entire vertebrate phylogeny. This is because the molecule plays a very important role in gas transport. Thus, changes in the genome frequently would cause deleterious changes in the molecule structure. This deleterious changes would lead to the elimination of this variant in the population. However, what Kimura presented is that nucleotide mutation occur at a very fast rate in the hemoglobin gene. These changes nearly never cause alterations in the translated molecule. Due to it, the researcher called them “Neutral mutations”. He also postulated that these mutations were the main substitutions which happen in the genome. Moreover, this changes happen in a previously unimaginable rapid rate, not only in the hemoglobin gene, but in the entire genome of all living organisms. This caused a huge change in scientific thinking of genome evolution and mutation, leading to the creation of models of phylogenetic inference and lots of other changes in other biological fields as well.

Correção José Serrano (31/03/17)

Good essay, although there are some points to review:

1. There is a lack of connection between the main idea of the essay “Stochastic changes” and the other ideas as history, importance of Kimura’s hypothesis, etc.
2. Try to avoid passive voice.
3. The word “this” is used without any subject even at the start of sentences. Example: “This caused a huge change in…”, that sentence is not connect with the previous one, what caused this?
4. The ideas before Kimura’s hypothesis should have at least one reference.
5. The central part of the essay could be synthesized to understand better all the Kimura’s ideas, for example: Kimura (1968) postulated the following points:…
6. This changes should be plural “these changes”.
7. Lots of other changes could be changed with many other changes.

Ensaio 04 (31/03/2017)

An example of direct influence in effective population size

Público-alvo: estudantes de graduação em Biologia.

Many intrinsic and extrinsic factors can influence a population’s effective size. The effective size of a population (expressed by Ne) is a value that “determines the rate of change in the composition of a population caused by genetic drift” (Charlesworth, 2009). In other words, this rate expresses how many individuals really contribute with alleles to the composition of the next generation. The Ne value of a population Hardy-Weinberg equilibrium is equal to the real population size. Thus, any disturbance in this equilibrium causes a reduction in the Ne. An example is a population with more females than males and that one male mates with one female. The number of individuals contributing to the next generation alleles (Ne value) is the number of copulating males and females. This number is less than the population size, given that some females would not have a mate. Therefore, these maleless females would not contribute to the next generation’s alleles, making the Ne smaller than the total population size. Besides this example, there are many other important events which reduce Ne, such as variation in offspring numbers, mode of inheritance, inbreeding and spatial structure. Therefore, knowing the effective population size is extremely important when infering something about it.

Comentários (Flávia)

Eu gostei muito do texto, do tamanho das frases, da fluidez e de uma frase final. Eu colocaria a sua segunda frase como frase tópico, porque ela realmente introduz o tema. Mudaria a frase final também, especificando um pouco mais (deixando menos no ar, mudando a frase "when infering something about it" por algo como "for studies in the area…". Muito bom

Ensaio 05 (05/05/2017)

Historical reconstruction as a tool not a goal.

Target audience: Biology students.

Systematists and other scientists who publish phylogenetic trees usually think this tree as their work’s final aim. This was possible to verify during the debate regarding this topic in class. What we discussed during class was that the historical reconstruction (= phylogenetic reconstruction/tree) is, in fact, a tool and a method to achieve final goals. These goals are not the historical reconstruction per se unless for those who work with development of new methods of phylogenetic reconstruction. For anyone else, the historical reconstructions are a way to achieve another goal. For example, a systematist uses the historical reconstruction to propose valid evolutionary classification for the group he/she is working with; a biogeographer uses the phylogenetic reconstruction to infer geological events which caused the divergence between groups of organisms; and a geneticist can use a phylogenetic tree to see from which group a specific gene came from. So, as users of this tool, people in class say that we should not give so much importance for the debate of those whose research is regarding the methods. We should rather only use it. I say that, if one do not follow this discussion, to know which methodology is epistemologically and logically better to use, becomes a very difficult topic. Therefore, what scientists end up doing is using what most of other scientists in their area do. I don’t blame who choose to do this way. After all, researchers have plenty of other more important issues to care about than if it is better to use parsimony or Bayesian in the phylogenetic inference. But if one wants not to “go with the flow” and using what everyone uses, he/she really should read the discussions and see which part he/she agrees with.

Correção José Serrano (12/05/17)

I think it is a good essay. I have no critics about the approach you used in your essay, although some considerations of drafting:

- The paragraph starting at "For anyone else…" is too long, you can easily cut it in 2 or 3 sentences.
- At the start of a new sentence try tu use a more formal conector. For example you could change But with However, So with Therefore and so on.
- Try to avoid abbreviations as don't for example

Once again goog essay

Cheers

Ensaio 06 (12/05/2017)

In a forest of trees as big as the universe, how to search for the best one?

Target audience: Biology students

The number of trees a computer program must search to present you the most parsimonious/ probable one is, very often, bigger than the number of atoms in the universe. Estimates in any physics website are that there are around 1078 to 1082 atoms in the known universe. In a phylogenetic analysis using three terminals, the number of possible rooted trees are only three. In an analysis using four terminals, this number grows to 15. If we use 30 terminals (which is not that much nowadays), there are 5x1038 possible trees. And if we increase the number of terminals to 80, there are around 3x10139 possible trees generated from the dataset. In other words, this number is bigger than the number of atoms in the known universe. Computationally speaking, this is an impossible number of trees for a computer to cover in a man's life. So, in the middle of this “forest” of trees, how can a program find the best one? There are two basic ways. One is the called exhaustive search, in which the computer analyzes all the possible trees and choose the best one, given the method the researcher has chosen. As we have just seen, this is not possible for most of the projects, once they deal with more than 20 terminals. The second and most used option is the heuristic search. In this search, the researcher limits the universe of trees that the program will search. One does so by introducing parameters of search. By doing this, the person limits the search for only those that are the most probable to be the best ones. The universe of trees one excludes from the analysis are those which are not worth checking, once they will not be the most probable anyway. Therefore, knowing these parameters used in a heuristic search and which one to employ in your analysis is one of the fundamental initial steps to begin the systematic research.

Comentário (Pietro)

Caio, gostei muito do seu texto e da forma como ele é fluido. Também achei ele adequado para o público alvo definido. Está de parabéns! A frase final está de acordo com o parágrafo apresentado, servindo como uma importante conclusão do assunto que vc quer falar. Entretanto a primeira sequência, apesar de trazer uma informação muito interessante não me pareceu uma sequência tópico, o que você fala no parágrafo vai além disso. Como seu ensaio possui um título acredito que isso possa estar minimizado, mas acho que pode ser válido pensar se teria como introduzir uma frase mais abrangente na abertura do ensaio.

Ensaio 07 (23/05/2017)

Bootstrap misemployed in phylogenetic analysis

Target audience: Systematics graduate students

Systematists generally interpret the bootstrap value wrongly. What we usually see in systematic works is researchers employing this statistic values to say that a specific node has a determined probability of being real. For example, if a node has a bootstrap value of 0.95, people generally interpret this value as the chance of the clade be a true clade in the nature. In other words, we could say with 95% sure that this clade really branched in that way. However, what the bootstrap value really means is how many times a node/clade appears in an analysis if some subtle removals are done in the dataset. After the removals, the algorithm runs the analysis again to check the clades recovered. So, if a clade appears 95 times out of 100 replicas, then the bootstrap value of this clade is 0.95. Then, we can see that this support value is not linked to the "reality" in any way. What it express, in fact, is how reliable are the clades taking that specific dataset into consideration. If a clade presents a low bootstrap value, it does not mean that it does not exist. What this low support value tells is that, given your dataset, the clades recovered in the best tree are not recovered in most of the analysis removing little amount of data. Therefore, bootstrap value indicates how feasible one's dataset is to perform the analysis he/she wants to do. If someone recovers a low support value, what one should think is not that the clades does not exist, but yes that more data should be collected to make the anaysis more robust.

Ensaio 08 (26/05/2017)

Parameters considered in a Maximum Likelihood analysis

Target audience: Systematics graduate students

There are four kinds of parameters considered to build a model in a Maximum Likelihood analysis (ML). One of them is the substitution rate from one nucleotide to another. In this parameter, there are costs related to each substitution (π). There is a continuum from a model in which all the substitutions have the same cost (e.g. a substitution from “A” to “T” has the same cost of one from “A” to “G”) until one in which each substitution has a different cost. Another parameter takes into account the proportion of each nucleotide (A, T, C and G) in a site. This is important because in an Archaea, for example, proportion of “C” and “G” is bigger than the proportion of the others. Therefore, each dataset will have its π parameter best suited to explain the changes. The third parameter is the search for invariable sites and exclude them of the substitution parameter. With this parameter, the computational time reduces. The last parameter estimates the variation of π in different sites. This is an important data to consider once the substitutions occur in different ways in different sites. For example, the second base of a codon varies less than the third base. Taking into account these differential substitutions is important to create a more realistic model. Understanding these parameters is a fundamental step to know which model best suits your data. On the contrary, you will let the “default” function of a program make it for you and probably the model will not fit into your dataset and your results will not reflect the best relationship of your taxa.

comentários: Lívia Moura
Olá Caio. Grata pela avaliação :)~~
Sobre seu texto, achei muito bom, adequado, claro e informativo ao público alvo. Você teve um cuidado com a forma escrita em inglês que não vi em muitos textos, e o parabenizo por isso. No primeiro instante, não havia achado sua frase tópico adequado, porém, quando li seu título, ela tornou-se perfeita.

Ensaio 09 (02/05/2017)

How phylogenetic algorithms explore “forrest of trees”

Target audience: Biology students

When using search algorithm in phylogenetic reconstruction,how does it search the most likely/parsimonious one? It is a very complicated matter even for a computer to solve, once the number of possible trees is impressively huge even for a small dataset (Felsenstein, 2004). What the algorithms generally do is to employ search strategy called Hill-Climbing (Giribet, 2007). In this search strategy, the algorithm starts to explore the possible trees in a specific region of the space of trees. If it finds a tree which is better than the previous one, he takes this as the best. This step is repeated exhaustively until the algorithm finds the tree that considered the “best” of that local. However, the tree that the algorithm considers the best is not necessarily the best of all trees. If the algorithm started exploring in another region, it could reach a different result. So, there are “tricks” to make the algorithm explore more than one region. These are steps added to the search algorithm to make it explore other regions after it reaches the best “local” result. After the algorithm finds the best local tree, these new steps make it start again from another region and explore the trees of this new local until it reaches the best one there. The most likely/ parsimonious is the one retained. In short, the search algorithms complemented with these refinements increases the chances of finding the best tree to explain one specific dataset.

Ensaio 10 (21/06/2017)

Aprendizados ao longo do curso

A presente disciplina é muito bem estruturada. Já na primeira aula percebi que o encadeamento das partes de cada aula e o encadeamento de todas as aulas foram muito bem pensados. Quando entrei, acreditava que poderia aprender mais aspectos instrumentais das análises filogenéticas, como o uso do softwares. Porém, na primeira aula já ficou claro que isso não ocorreria e me senti mais entusiasmado ainda para cursar a disciplina e discutir aspectos mais profundos do que apenas ferramentas de análise filogenética. Acredito que as discussões foram extremamente produtivas para que eu olhasse a inferência filogenética molecular de outro ponto de vista. Quando entrei na disciplina, acreditava que as técnicas moleculares eram algo simples de serem empregadas e de forma muito exata. Ao final da disciplina, percebo que isso não é nem um pouco verdade e que a complexidade dos dados moleculares excede e muito a complexidade dos caracteres morfológicos.
Além do aprendizado de conteúdo, desenvolvi a habilidade de produção textual. Li o meu primeiro texto e comparei com meu último e percebi uma melhoria muito grande na forma em que estou escrevendo. E a escrita dessa forma que faz parte de minha praxis agora. Penso nos aspectos de uma boa escrita mencionados em aula e presentes no "livrinho" todas as vezes que preciso escrever textos. Isso era uma dificuldade particular muito grande, dado que um artigo meu ficou seis meses sem ser publicado devido à escrita pouco clara e não científica.
Mesmo tendo aproveitado muio bem a disciplina, acredito que poderia ter aproveitado mais dois aspectos. Um deles é o das leituras das aulas. Enquanto a leitura estava sendo prévia, me planejei para me manter em dia com os textos. Quando a leitura mudou para póstuma à aula, li apenas dois textos. Isso porque me perdi em meu próprio planejamento em relação às aulas. Um ouro aspecto foi o de que faltei na última aula e acredito que ela teria sido fundamental para à aplicação de métodos de inferência filogenética em meu mestrado.
Com a somatória de esforço para produção de bons textos, atenção na aulas, empenho na produção do seminário e a combinação falta de mais esforço em leitura (mas não muito) e ausência em uma aula importante, creio que a nota de auto-avaliação justa para mim é 0,85.

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