Update: I’m currently taking MIT’s 7.00x Introduction to biology course online at Edx.org. Highly recommended.
Hello friends! Unfortunately I will have to put this biology learning project on hold. I’m currently focussing my efforts on a new startup: Biohack Stack.
There’s never enough time to do everything I want. One lifetime is not enough! Reversing ageing will address this problem.
I’m attempting to compress a 4-year biology degree into 1 year of self-directed study so I can eventually launch a startup in biotech / anti-aging. I’m starting off with a basic overview of Molecular Biology of the Cell by Alberts et al. (textbook text here) Here are my personal thoughts and questions from Chapter 1.[[Biology]]
- Hereditary information of organisms is encoded in DNA.
- The alphabet of DNA consists of four nucleotide monors: adenine, guanine, thymine, and cytosine.
- The backbone of DNA is formed from a chain of repeating sugar and phosphate.
- Nucleotides pair up in a complementary way. C-G, A-T. There are hydrogen bonds in between the nucleotides. These weak bonds allow the DNA helix to be pulled apart without break individual DNA strands.
- The two strand nature of DNA facilitates replication. Split the two strands and then each strand can fill out with its complementary nucleotides. This is called templated polymerization.
- Question: What function does the twist of the DNA helix play? What force causes this twist?
- DNA –> tRNA –> amino acids –> proteins
- RNA is similar to DNA but with a slightly different sugar backbone. Also the thymine monomer is replaced with Uracil. WHY? I hope there is a satisfying reason why only one of the nucleotides are replaced going from DNA –> RNA.
- Three letters of RNA (U, A, C, G) form a codon which codes for an amino acid. There are 64 combinations but only 20 amino acids. i.e degeneracy.
- Not all genes in our DNA code for proteins. Some genes code for RNA that have catalytic, regulatory, or structural function.
- The same gene can create alternative versions of proteins depending on the way the RNA is processed.
- The human genome has 3 billion base pairs. About 30,000 genes.
- 60 genes are common to all life on Earth.
- About 200 genes can be found in all 3 major groups of life on earth: Bacteria, Archaea, and Eukaryotes.
- Prokaryotes do not have a membrane enclosing their DNA. Eukaryotes do.
- Archaea are a bit like bacteria but also a but like eukaryotes.
- The simplest bacteria have about 500 genes.
- Eukaryotes have about an order of magnitude more genes than prokaryotes and about 3 orders of magnitude more non-coding DNA. This is because multicellular organisms DNA has more complex regulation.
- New genes are formed from old genes. It’s not randomly spelt one letter at a time.
- There are four ways to make new genes: Intragenic mutation (an error in a base pair during replication), duplication of a gene, DNA segment shuffling (hybridization), or horizontal gene transfer.
- Duplication of genes occurred a lot in the evolution of vertebrates.
- Horizontal gene transfer happens a lot in bacteria. Phages can inject new DNA into bacteria. Bacteria can also take in DNA / RNA segments floating around in the environment. Anti-bacterial resistance can be conferred through horizontal gene transfer.
- New genes formed from old genes can be related and have similar function. We can group genes into families.
- Genes that are related but occur in two divergent species are called orthologs.
- Genes that are related but occur in the same species are called paralogs.
- Related genes (ortho and para) are called homologs.
- Often we can get clues about functions of novel genes by comparing them to a database of existing genes.
- Homolog genes can span across yeast –> humans. Replacing some yeast genes with a human homolog gene may result in a fully functional yeast organism. CRAZY.
- We can determine the function of a gene through cell biology / observation of mutants and also biochemistry.
- The author(s) are very optimistic that the complexity of biology can be overcome and understood at the cellular level, especially with the use of modern day computing. I wonder what the hard limits are.
- Modelling something like an entire cell seems intractable but perhaps there are some approximations that can be made.
- What is the most complex system we have managed to “accurately” (need qualifications here) model? How many orders of magnitude more complex is a cell?
Crazy Tangents: Physics, Ergodic Theorem, and Nassim Taleb
- There seems to be some major theme of non-ergodicity in evolutionary biology / genetics. I’ve been thinking a lot about ergodicity ever since I read Nassim Taleb’s Skin in The Game.
- Taleb believes much of financial modelling is incorrect because it assumes ergodicity; that the randomness of financial markets can be modelled as a random sample of its phase space (possible states).
- In real life ergodicity is broken because if you go bust on day 10, there will be no day 11. You can think of think of going bust as an absorbent barrier.
- In biology, mutations are “random” yet we get increasing complexity. Natural selection enforces non-ergodicity with its own absorbent barriers– pruning “bad” mutations and favouring the good ones.
- Ergodicity features prominently in statistical mechanics and many of the fundamental results in physics rest on this assumption. The models we have built on top of this assumption are accurate as far as we know.
- But what if ergodicity is only approximately true in physics? What sort of absorbent barriers could be postulated in a Boltzman model context? I’m guessing none. We probably just need to recognize we very rarely encounter a box of gas that does not interact with it’s environment in real life.
- Thinking about the uber long term game of longevity, statistical mechanics and entropy (2nd Law of Thermodynamics) is going to be a major concern.
- Throwing it out there: Perhaps there is a link to the information blackhole paradox in physics. Blackholes could act as an absorbent barrier that breaks the ergodic theorem on some fundamental level.
- Just so happens I stumbled on this paper in the comments section of Hacker News Today: Statistical Physics of Adaptation. https://arxiv.org/pdf/1412.1875.pdf
1-1. F. The human haemoglobin genes within the same family in the same species (homo sapien) are considered to be paralogs. Orthologs are related genes across different species that diverged evolutionarily.
1-2. T. Horizontal gene transfer is easier to do in unicellular organisms that only have one strand of DNA to modify. Horizontal gene transfer is possible in multicellular organisms but the transfer would need to affect germ line cells in order for the change to be passed on in the genome of the progeny.
1-3. T. Most bacteria DNA is coding DNA. Most eukaryote DNA is non-coding.
1-4. Natural selection has selected for organisms that have DNA error-correction / error resistance. Organisms without such ability would be at a huge disadvantage in propagating their genes.
1-5. Sequence the DNA to see if the genes are related to any genes of organisms on Earth. If there are similarities it is either contamination or somehow life evolved in a similar manner on Europe — possible but unlikely.
1-6. Feeding is merely energy extraction. Photons from the sun have energy. Mixtures of chemicals and ions can also have stored energy.
1-8. Genes for ribosomal RNA were not necessarily “born” “perfect”. In fact, it is highly unlikely. The genes evolved slowly but that does not mean they were the best possible genes. This question has a lot of imprecise language.
1-9. A complex process probably has more “moving parts” and proteins / RNA involved. The addition of horizontal gene transfer may have a greater probability of disrupting the process compared to one that is simpler and has fewer proteins / RNA.