Bilingual

Is biology too complex to ever understand? - Bert Hubert's writings
a day ago
- Biology historically lacked mathematical rigor, relying on observation and often incorrect classifications.
- Mendel's pea experiments introduced quantitative laws, attracting physicists to biology.
- Physics achieved unification through simple formulas and particles, inspiring 'physics envy' in biology.
- Initial molecular biology successes (DNA, codons, insulin production) mirrored physics' reductionism.
- Despite progress, most biological systems (e.g., lac operon, diabetes, Alzheimer's) remain poorly understood due to immense complexity.
- Biology produces over 100 articles per hour, exceeding human capacity for synthesis.
- Alternative approaches include focusing on 'enough' understanding for practical cures (e.g., hepatitis C) and prioritizing machine-readable databases over traditional papers.
- Fundamental research on universal aspects of life (e.g., idealized DNA) may yield deeper unified principles.
DNA seen through the eyes of a coder (or, If you are a hammer, everything looks like a nail) - Bert Hubert's writings
a day ago
- DNA is like byte-compiled code for a virtual machine called the nucleus, using a digital but non-binary language with four bases (T, C, G, A) and three-letter codons that encode amino acids.
- Transposable elements (jumping DNA) are analogous to position-independent code, allowing stretches of DNA to move around the genome.
- Conditional compilation (#ifdef) is seen in how cells express only relevant genes, with stem cells being like un-#ifdefed code that can differentiate into any cell type.
- Epigenetics and imprinting act like runtime binary patching, modifying gene activation without altering DNA sequence, and can be heritable.
- Junk DNA (introns) functions like comments that are physically spliced out, and may aid in DNA folding, though its exact role remains debated.
- Cell division (fork()) can lead to tumors (fork bombs) when safeguards fail, similar to the halting problem in computing.
- The genetic code uses degenerate encoding (Reed-Solomon-like) to protect against errors, and highly conserved genes (like histones) are 'holy code' that rarely mutate.
SkewDB - Bert Hubert's writings
a day ago
- SkewDB is a free database containing GC and other skew data for over 53,000 chromosomes and plasmids, last updated on July 5, 2024, now including homopolymer counts and codon bias data.
- The database is created using open-source Antonie software, based on public DNA sequence, annotation, and taxonomic data from NCBI, and provides individual files with 4096-nucleotide granularity for each chromosome.
- Key features include fields for skew gradients (e.g., GC, TA, strand bias), codon position analysis, taxonomic classification, and fit quality metrics, with the project aiming to support research on GC skew and chromosome behavior.

Hasty Briefsbeta

bioinformatics

Is biology too complex to ever understand? - Bert Hubert's writings

DNA seen through the eyes of a coder (or, If you are a hammer, everything looks like a nail) - Bert Hubert's writings

SkewDB - Bert Hubert's writings