Chapter 1
Foundations & Databases
1: Biological Databases, Data Schemas & Accession Ontologies
- • Primary vs derivative vs curated databases: GenBank, RefSeq, and UniProtKB
- • Data schemas and flat-file anatomy: FASTA, FASTQ, GenBank, GFF3, and mmCIF
- • Accession number versioning and stability across biological revisions
- • Biomedical ontologies: Gene Ontology (GO) Directed Acyclic Graphs (DAGs) and Sequence Ontology (SO)
- • Programmatic API querying using NCBI E-utilities (esearch, efetch) and Ensembl REST
Chapter 2
Sequence Alignment
2: Global Sequence Alignment: Needleman-Wunsch Algorithm
- • Mathematical definition of pairwise sequence alignment: Homology and evolutionary distance
- • The Needleman-Wunsch dynamic programming recurrence relation for global alignment
- • Step-by-step matrix initialization, fill, and pointer array traceback
- • Gap penalties: Constant vs linear vs affine gap models ($W = -d - (k-1)e$)
- • Time and space complexity analysis ($O(mn)$) and biological applicability
Chapter 3
Sequence Alignment
3: Local Sequence Alignment: Smith-Waterman Algorithm
- • Why global alignment fails for multidomain proteins and disparate gene lengths
- • The Smith-Waterman recurrence relation with non-negative zero threshold
- • Matrix initialization: Initializing row 0 and column 0 to zero without gap penalties
- • Locating the global maximum cell and terminating traceback at zero
- • Space-saving divide-and-conquer: Hirschberg's linear space $O(m)$ algorithm
Chapter 4
Evolutionary Models
4: Scoring Matrices: PAM, BLOSUM & Log-Odds Mathematics
- • Physicochemical properties of amino acids and conservative vs radical substitutions
- • Margaret Dayhoff's Point Accepted Mutation (PAM) Markov transition model
- • Steven and Jorja Henikoff's Block Substitution Matrix (BLOSUM) methodology
- • Log-odds scoring mathematics: Converting substitution probabilities into additive half-bit units
- • Selection criteria: When to choose BLOSUM62 vs BLOSUM80 vs PAM250 for target evolutionary distances
Chapter 5
Database Searching
5: Heuristic Sequence Searching: BLAST & FASTA Algorithms
- • Why dynamic programming is computationally intractable for billion-sequence database searches
- • The BLAST heuristic architecture: Seed generation, neighborhood lookup, and ungapped extension
- • Karlin-Altschul extreme value distribution (Gumbel) statistics: Normalized bit score ($S'$) and $E$-value
- • The 5 standard BLAST programs: blastn, blastp, blastx, tblastn, and tblastx
- • Command-line BLAST+ execution, custom database creation (makeblastdb), and tabular parsing
Chapter 6
Sequence Alignment
6: Multiple Sequence Alignment (MSA) & Progressive Methods
- • NP-hardness of exact multidimensional dynamic programming: $O(L^N)$ complexity
- • Progressive alignment heuristic: Pairwise distance calculation, guide tree clustering, and profile alignment
- • The 'Once a gap, always a gap' flaw in ClustalW and frozen error propagation
- • Iterative refinement algorithms: MUSCLE log-expectation scoring and profile-profile reassessment
- • Ultra-scale alignments using MAFFT FFT (Fast Fourier Transform) heuristics for viral and bacterial cohorts
Chapter 7
Phylogenetics
7: Molecular Phylogenetics & Evolutionary Tree Reconstruction
- • Anatomy of phylogenetic trees: Clades, nodes, branch lengths, rooted vs unrooted topologies
- • Distance matrix methods: UPGMA (ultrametric clock assumption) vs Saitou-Nei Neighbor-Joining (NJ)
- • Nucleotide substitution models: Jukes-Cantor (JC69), Kimura 2-parameter (K80), HKY85, and General Time Reversible (GTR)
- • Maximum Likelihood tree reconstruction and ModelFinder with IQ-TREE
- • Assessing node reliability: Non-parametric Felsenstein bootstrapping and UFBoot
Chapter 8
Regulatory Genomics
8: Sequence Motifs, Position Weight Matrices & Information Content
- • Biological definition of motifs: Transcription factor binding sites (TFBS) and splice junctions
- • Representing degenerate binding sites: IUPAC ambiguity codes vs Position Frequency Matrices (PFM)
- • Calculating Position Weight Matrices (PWM) using log-odds likelihood and pseudocounts
- • Information content theory: Claude Shannon entropy and letter height in bits (maximum 2 bits for DNA)
- • Scanning genomes for regulatory motifs using MEME-Suite (FIMO) and BioPython
Chapter 9
Probabilistic Modeling
9: Hidden Markov Models (HMMs) in Biological Sequence Analysis
- • Mathematical definition: Visible sequence observations vs hidden biological states
- • The Three Foundational HMM Problems: Evaluation (Forward), Decoding (Viterbi), and Learning (Baum-Welch)
- • The Viterbi dynamic programming algorithm for most probable state path reconstruction
- • Profile HMMs: Modeling protein families with match, insertion, and deletion states
- • Searching protein domain databases (Pfam / InterPro) with HMMER3 (hmmsearch, hmmscan)
Chapter 10
Statistical Modeling
10: Quantitative Transcriptomics & Statistical Differential Expression
- • Why normal and Poisson distributions fail for RNA-seq: Biological overdispersion
- • The Negative Binomial distribution: Mean-variance relationship ($Var(Y) = \mu + \alpha \mu^2$)
- • Empirical Bayes shrinkage: Information borrowing across thousands of genes in DESeq2 and edgeR
- • Hypothesis testing: Wald test for pairwise contrasts vs Likelihood Ratio Test (LRT) for multi-factor designs
- • Multiple testing correction: Controlling the False Discovery Rate (FDR) with Benjamini-Hochberg
Chapter 11
Structural Biology & AI
11: Protein Tertiary Structure Prediction & The AlphaFold Era
- • The 50-year protein folding challenge: Levinthal's paradox and thermodynamic hypothesis
- • Homology modeling vs ab initio Rosetta vs deep learning breakthroughs
- • AlphaFold 2/3 architecture: Multiple Sequence Alignment (MSA) representation and Pair Representation
- • The Evoformer block and Invariant Point Attention (IPA) 3D coordinate frame refinement
- • Interpreting structural confidence: Local pLDDT per residue and global Predicted Aligned Error (PAE) matrices
Chapter 12
Systems Biology
12: Network Biology, Protein-Protein Interactions & Graph Theory
- • Graph theoretical foundations: Nodes, edges, adjacency matrices, and directed vs undirected graphs
- • Topological properties: Degree distribution, clustering coefficient, path length, and betweenness centrality
- • Biological networks as Scale-Free graphs: Barabasi-Albert preferential attachment and lethal hub proteins
- • Protein-Protein Interaction (PPI) networks: Querying the STRING database programmatically
- • Community detection algorithms: Markov Clustering (MCL) and Louvain modularity optimization
Chapter 13
AI & Machine Learning
13: Machine Learning & Deep Learning in Modern Genomics
- • Formulating genomic problems as machine learning tasks: Classification, regression, and generative modeling
- • Feature engineering on biological sequences: One-hot encoding, $k$-mer frequency vectors, and conservation profiles
- • Convolutional Neural Networks (1D CNNs) for learning transcription factor binding motifs directly from DNA (DeepBind architecture)
- • Genomic Foundation Models: Transformer architectures trained on unaligned genomes (DNABERT, Nucleotide Transformer, HyenaDNA)
- • Rigorous evaluation: Chromosome-split cross-validation to prevent data leakage and AUROC/AUPRC assessment
Peer-Reviewed Foundations
Seminal peer-reviewed benchmarks and methodology literature
📚 Foundational Literature & Methodology Benchmarks
Foundational Sequence Algorithms
- Needleman SB, Wunsch CD. (1970). A general method applicable to the search for similarities in the amino acid sequence of two proteins. J Mol Biol 48(3):443–453. doi:10.1016/0022-2836(70)90057-4
- Smith TF, Waterman MS. (1981). Identification of common molecular subsequences. J Mol Biol 147(1):195–197. doi:10.1016/0022-2836(81)90087-5
- Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. (1990). Basic local alignment search tool (BLAST). J Mol Biol 215(3):403–410. doi:10.1016/S0022-2836(05)80360-2
Probabilistic Models & HMMs
- Eddy SR. (2011). Accelerated profile HMM searches (HMMER3). PLOS Comput Biol 7(10):e1002195. doi:10.1371/journal.pcbi.1002195
- Sievers F, Wilm A, Dineen D, et al. (2011). Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol 7:539. doi:10.1038/msb.2011.75
- Felsenstein J. (1981). Evolutionary trees from DNA sequences: a maximum likelihood approach. J Mol Evol 17(6):368–376. doi:10.1007/BF01734359
Structural AI & Systems Biology
- Jumper J, Evans R, Pritzel A, et al. (2021). Highly accurate protein structure prediction with AlphaFold. Nature 596:583–589. doi:10.1038/s41586-021-03819-2
- Shannon P, Markiel A, Ozier O, et al. (2003). Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 13(11):2498–2504. doi:10.1101/gr.1239303
- Nguyen E, Poli M, Faizi M, et al. (2024). Sequence modeling and design from molecular to genome scale with HyenaDNA. Nat Commun 15:7001. doi:10.1038/s41467-024-51000-y