1: Biological Databases, Data Schemas & Accession Ontologies

Chapter 1: Biological Databases, Data Schemas & Accession Ontologies

Modern bioinformatics is founded upon structured, publicly accessible repositories of biological sequence, structure, and functional data. Navigating these databases and parsing their underlying schemas programmatically is the indispensable initial capability required of every computational biologist.

1. The Global Biological Database Ecosystem

• NCBI Entrez (National Center for Biotechnology Information): Federates PubMed, Nucleotide, Protein, Gene, ClinVar, and dbSNP through a unified text query engine.
• Ensembl & GENCODE: Provides high-quality, comprehensive reference gene annotations for vertebrates, rigorously defining canonical transcripts, exons, and splice variants.
• Protein Data Bank (PDB): The worldwide single archive of experimentally determined 3D structures of biological macromolecules determined via X-ray crystallography, NMR, and cryo-EM.

2. Gene Ontology: Directed Acyclic Graphs (DAGs)

Free-text gene descriptions make automated computation impossible. The Gene Ontology (GO) provides a controlled, standardized vocabulary organized as a Directed Acyclic Graph across three orthologous domains:

• Molecular Function (MF): Elemental biochemical activities of a gene product at the molecular level (e.g., ATP-dependent kinase activity, GO:0016301).
• Biological Process (BP): Coordinated series of molecular events or biological pathways (e.g., glycolytic process, GO:0006096).
• Cellular Component (CC): Subcellular anatomical structures and macromolecular complexes (e.g., mitochondrial inner membrane, GO:0005743).

3. Programmatic Database Retrieval with Python

# Query NCBI Nucleotide programmatically via Entrez E-utilities
import urllib.request
import xml.etree.ElementTree as ET

# 1. Search for human BRCA1 mRNA transcripts
search_url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=nuccore&term=BRCA1[Gene]+AND+Homo+sapiens[Organism]+AND+refseq[filter]&retmode=json"
with urllib.request.urlopen(search_url) as resp:
    import json
    search_data = json.loads(resp.read().decode())
    uid_list = search_data['esearchresult']['idlist']
    print(f"Discovered {len(uid_list)} RefSeq UIDs: {uid_list[:3]}")

# 2. Fetch the FASTA record for the primary hit
fetch_url = f"https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=nuccore&id={uid_list[0]}&rettype=fasta&retmode=text"
with urllib.request.urlopen(fetch_url) as resp:
    fasta_text = resp.read().decode()
    print("FASTA Header:", fasta_text.splitlines()[0])