Day 1: Biological Databases & Programmatic Sequence Retrieval (NCBI, UniProt, PDB)

1. Foundations of Biological Databases in Molecular Biology

Modern computational molecular biology rests upon centralized, standardized, and curated biological data repositories. High-throughput automated sequencing and structural biology consortia deposit gigabytes of sequence and coordinate information daily. To extract biological insight, researchers must understand the foundational taxonomy of biological repositories:

2. The FASTA Sequence Format Standard

The universal currency of sequence analysis is the FASTA format. A standard FASTA file consists of a single-line definition header starting with a greater-than symbol (>), followed immediately by a unique sequence identifier and arbitrary descriptive metadata, followed by lines of nucleotide or amino acid codes using IUPAC standard nomenclature:

>NM_000546.6 Homo sapiens tumor protein p53 (TP53), transcript variant 1, mRNA
GATGGGATTGGGGTTTTCCCCTCCCATGTGCTCAAGACTGGCGCTAAAAGTTTTGAGCTTCTCAAAAGTC
TAGTCCAAAAATAATTTCATCACTTTTTTTTTTTTTTTTTAATTTTTTTTTTTTTTTTTTTTTTTTTTTT
TTTTTTTTTTTTTTTGTCTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT
...

IUPAC Degenerate Nucleotide & Amino Acid Codes

When sequencing primers, degeneracies, or unknown bases are present, computational tools rely on standard IUPAC ambiguous nucleotide codes:

3. Programmatic Sequence Retrieval with NCBI Entrez E-utilities

NCBI's Entrez E-utilities (Entrez Programming Utilities) represent a suite of server-side REST APIs providing programmatic access to all NCBI databases. The core utility pipeline involves:

• esearch: Queries an NCBI database with Boolean logic and returns a list of matching Primary Identification (UID) numbers.
• efetch: Downloads complete biological records (FASTA, GenBank XML, ASN.1) for specified UIDs.
• esummary: Returns brief document summaries (DocSums) without downloading complete sequences.
• elink: Traverses cross-database relationships (e.g. linking a gene ID in gene to its related structures in structure or articles in pubmed).

Production Biopython Script: Automated Multi-Database Sequence Fetcher

#!/usr/bin/env python3
"""
Day 1 Practical: Programmatic Retrieval of Genes, Transcripts, and Proteins
Integrates NCBI Entrez, UniProt REST, and RCSB PDB APIs.
"""

import sys
import json
import urllib.request
from Bio import Entrez, SeqIO

Entrez.email = "bioinformatics.researcher@omics.co.in"
Entrez.tool = "OmicsPortalResearch"

def fetch_ncbi_nucleotide(accession: str, rettype="fasta"):
    """Fetch nucleotide record from NCBI Nucleotide/RefSeq."""
    print(f"[NCBI] Retrieving {accession} (type: {rettype})...")
    try:
        handle = Entrez.efetch(db="nuccore", id=accession, rettype=rettype, retmode="text")
        record_text = handle.read()
        handle.close()
        return record_text
    except Exception as e:
        print(f"Error fetching NCBI record {accession}: {e}", file=sys.stderr)
        return None

def fetch_uniprot_entry(uniprot_id: str):
    """Retrieve curated Swiss-Prot annotations via UniProt REST API."""
    url = f"https://rest.uniprot.org/uniprotkb/{uniprot_id}.json"
    print(f"[UniProt] Querying curated JSON for {uniprot_id}...")
    req = urllib.request.Request(url, headers={"User-Agent": "BiopythonWorkflow/1.0"})
    with urllib.request.urlopen(req) as response:
        if response.status == 200:
            data = json.loads(response.read().decode())
            gene_name = data.get("genes", [{}])[0].get("geneName", {}).get("value", "N/A")
            organism = data.get("organism", {}).get("scientificName", "N/A")
            function = "N/A"
            for comment in data.get("comments", []):
                if comment.get("commentType") == "FUNCTION":
                    function = comment.get("texts", [{}])[0].get("value", "N/A")
                    break
            seq = data.get("sequence", {}).get("value", "")
            return {
                "accession": uniprot_id,
                "gene": gene_name,
                "organism": organism,
                "function": function,
                "length": len(seq),
                "sequence": seq
            }
    return None

def fetch_pdb_structure(pdb_id: str, out_file: str):
    """Download macromolecular 3D coordinate file from RCSB PDB."""
    url = f"https://files.rcsb.org/download/{pdb_id.upper()}.cif"
    print(f"[RCSB PDB] Downloading mmCIF coordinates for {pdb_id.upper()}...")
    urllib.request.urlretrieve(url, out_file)
    print(f"Saved {pdb_id.upper()} mmCIF coordinates to {out_file}.")

if __name__ == "__main__":
    # 1. Fetch Human TP53 RefSeq mRNA
    tp53_fasta = fetch_ncbi_nucleotide("NM_000546.6", rettype="fasta")
    with open("TP53_human_transcript.fasta", "w") as f:
        f.write(tp53_fasta)
    print("Saved TP53_human_transcript.fasta")

    # 2. Fetch Human Insulin from UniProt (P01308)
    insulin_data = fetch_uniprot_entry("P01308")
    print(f"Insulin Gene: {insulin_data['gene']}, Organism: {insulin_data['organism']}")
    print(f"Sequence Length: {insulin_data['length']} amino acids")
    print(f"First 30 AA: {insulin_data['sequence'][:30]}...")

    # 3. Download PDB 1TUP (p53 core domain bound to DNA)
    fetch_pdb_structure("1tup", "1tup.cif")

4. Hands-On Laboratory Exercise

1. Query NCBI using esearch for the human BRCA1 gene (UID search: BRCA1[Gene Name] AND human[Organism] AND RefSeq[Filter]).
2. Download the canonical mRNA transcript (NM_007294.4) and canonical protein (NP_009225.1) using the automated script.
3. Inspect the FASTA header and verify the CDS coordinates in the corresponding GenBank flat file.