Theory2 Bioinformatics

Analyze DNA, RNA, protein sequences, and 3D protein structures using Biopython.

Sequence Analysis

Analyze Sequence

Analyze DNA, RNA, or protein sequences for basic properties:

# DNA sequence analysis
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics analyze-sequence \
  --sequence="ATGCGATCGATCG"

# Protein sequence analysis (auto-detected)
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics analyze-sequence \
  --sequence="MVLSPADKTNVK" --seq-type=protein --explain

# RNA sequence
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics analyze-sequence \
  --sequence="AUGCGAUCGAUCG" --seq-type=rna

Parameters:

• --sequence: DNA/RNA/Protein sequence string (required)
• --seq-type: auto, dna, rna, or protein (default: auto)
• --explain: Include detailed explanation
• --json: Output as JSON

Returns:

• Sequence type (DNA, RNA, or protein)
• Length in residues/nucleotides
• Composition (base/amino acid counts)
• Molecular weight (Da)
• GC content (for DNA/RNA only)

Translate DNA

Translate DNA sequences to protein using genetic code tables:

# Standard genetic code (table 1)
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics translate-dna \
  --sequence="ATGGCTAGCTAG"

# Mitochondrial genetic code (table 2)
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics translate-dna \
  --sequence="ATGGCTAGCTAG" --table=2

Parameters:

• --sequence: DNA sequence to translate (required)
• --table: Codon table (1=Standard, 2=Mitochondrial, default: 1)
• --json: Output as JSON

Returns:

• Original DNA sequence
• Translated protein sequence
• Amino acid count
• Codon table used

Find ORFs

Find Open Reading Frames (ORFs) in DNA sequences:

# Find ORFs with default 100bp minimum
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics find-orfs \
  --sequence="ATGAAATAG..." --min-length=100

# Find shorter ORFs (50bp minimum)
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics find-orfs \
  --sequence="ATGAAATAG..." --min-length=50

Parameters:

• --sequence: DNA sequence to search (required)
• --min-length: Minimum ORF length in nucleotides (default: 100)
• --json: Output as JSON

Returns:

• Number of ORFs found
• ORF list (up to 20) with:
  • Start position
  • End position
  • Length
  • Frame
  • Sequence

Protein Analysis

Analyze Protein

Compute detailed protein properties:

# Full protein analysis with explanation
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics analyze-protein \
  --sequence="MVLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHFDLSH" \
  --explain

Parameters:

• --sequence: Protein sequence (one-letter amino acid codes, required)
• --explain: Include detailed explanation
• --json: Output as JSON

Returns:

• Length (amino acids)
• Molecular weight (Da)
• Isoelectric point (pI)
• Instability index (<40 = stable)
• GRAVY score (hydropathicity)
• Aromaticity
• Amino acid composition (%)

Interpretation:

• Instability index < 40: Protein is stable in test tube
• GRAVY > 0: Hydrophobic protein
• GRAVY < 0: Hydrophilic protein

Find Domains

Search for protein domains/motifs using PROSITE-like patterns:

# Search for common domains
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics find-domains \
  --sequence="MVLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHFDLSH"

Parameters:

• --sequence: Protein sequence to search (required)
• --json: Output as JSON

Returns:

• Number of domains found
• Domain list with:
  • Domain ID
  • Domain name (e.g., N-glycosylation site, PKC phosphorylation)
  • Start position
  • End position
  • Sequence
  • Score/confidence

Searches for:

• N-glycosylation sites
• PKC phosphorylation sites
• Casein kinase II phosphorylation
• Myristoylation sites
• And other PROSITE patterns

Align Sequences

Perform pairwise sequence alignment:

# Global alignment (Needleman-Wunsch)
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics align-sequences \
  --seq1="ACGT" --seq2="AGCT" --alignment-type=global

# Local alignment (Smith-Waterman)
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics align-sequences \
  --seq1="ACGTACGTACGT" --seq2="CGTACGTA" --alignment-type=local

Parameters:

• --seq1: First sequence (required)
• --seq2: Second sequence (required)
• --alignment-type: global or local (default: global)
• --json: Output as JSON

Returns:

• Alignment type
• Alignment score
• Sequence identity (%)
• Aligned sequence 1 (with gaps)
• Aligned sequence 2 (with gaps)
• Number of gaps

Alignment Types:

• Global: Needleman-Wunsch - aligns entire sequences
• Local: Smith-Waterman - finds best local match

Structure Analysis

Load Structure

Load and analyze PDB structure files:

# Load PDB structure
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics load-structure \
  --pdb-file="/path/to/1abc.pdb"

Parameters:

• --pdb-file: Path to PDB structure file (required)
• --json: Output as JSON

Returns:

• PDB ID
• Number of models
• Number of chains
• Number of residues
• Number of atoms
• Chain IDs
• Resolution (if available)

Find Contacts

Find atomic contacts within a structure:

# Find contacts within 4.0 Å
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics find-contacts \
  --pdb-file="/path/to/1abc.pdb" --cutoff=4.0

# Find contacts in specific chain
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics find-contacts \
  --pdb-file="/path/to/1abc.pdb" --cutoff=5.0 --chain=A

Parameters:

• --pdb-file: Path to PDB structure file (required)
• --cutoff: Distance cutoff in Angstroms (default: 4.0)
• --chain: Filter by chain ID (optional)
• --json: Output as JSON

Returns:

• Number of contacts
• Cutoff distance used
• Contact list (up to 50) with:
  • Residue 1 details (name, number, chain)
  • Residue 2 details
  • Atom 1 name
  • Atom 2 name
  • Distance (Å)

Analyze Binding Site

Analyze residues around a potential binding site:

# Analyze 8Å radius around residue 100 in chain A
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics analyze-binding-site \
  --pdb-file="/path/to/1abc.pdb" --center-residue=100 --chain=A --radius=8.0

# Larger binding site (10Å)
/home/mikeb/theory2/.venv/bin/theory --json bioinformatics analyze-binding-site \
  --pdb-file="/path/to/1abc.pdb" --center-residue=150 --chain=B --radius=10.0

Parameters:

• --pdb-file: Path to PDB structure file (required)
• --center-residue: Central residue number (required)
• --chain: Chain ID (default: A)
• --radius: Search radius in Angstroms (default: 8.0)
• --json: Output as JSON

Returns:

• Center residue details
• Number of nearby residues
• Nearby residue list with:
  • Residue name
  • Residue number
  • Chain ID
  • Distance from center (Å)
  • Residue type classification

Expected Output Format

All commands return structured JSON:

{
  "status": "success",
  "result": {
    // Command-specific results
  },
  "metadata": {
    "tool": "Biopython",
    "method": "Analysis method",
    "timestamp": "ISO-8601",
    "duration_ms": 123,
    "gpu_used": false
  },
  "provenance": {
    "method": "Description",
    "inputs": {...},
    "library": "Bio.X.Y"
  },
  "next_actions": [
    "Suggested next steps"
  ]
}

Dependencies

All bioinformatics commands require Biopython:

# Install Biopython
uv pip install biopython

Common Workflows

Gene Discovery Pipeline

# 1. Find ORFs in genomic DNA
theory --json bioinformatics find-orfs \
  --sequence="..." --min-length=300

# 2. Translate ORF to protein
theory --json bioinformatics translate-dna \
  --sequence="ORF_SEQUENCE"

# 3. Analyze protein properties
theory --json bioinformatics analyze-protein \
  --sequence="TRANSLATED_PROTEIN" --explain

# 4. Find functional domains
theory --json bioinformatics find-domains \
  --sequence="TRANSLATED_PROTEIN"

Sequence Comparison

# 1. Analyze both sequences
theory --json bioinformatics analyze-sequence --sequence="SEQ1"
theory --json bioinformatics analyze-sequence --sequence="SEQ2"

# 2. Align sequences
theory --json bioinformatics align-sequences \
  --seq1="SEQ1" --seq2="SEQ2" --alignment-type=global

Structure-Function Analysis

# 1. Load structure
theory --json bioinformatics load-structure --pdb-file="protein.pdb"

# 2. Identify binding site
theory --json bioinformatics analyze-binding-site \
  --pdb-file="protein.pdb" --center-residue=100 --radius=10

# 3. Find contacts within binding site
theory --json bioinformatics find-contacts \
  --pdb-file="protein.pdb" --cutoff=5.0 --chain=A

Python API

from bioinformatics.sequence_analysis import (
    analyze_sequence,
    translate_dna,
    find_orfs,
    align_sequences
)
from bioinformatics.protein_analysis import (
    analyze_protein,
    find_domains
)
from bioinformatics.structure_analysis import (
    load_pdb_structure,
    find_contacts,
    analyze_binding_site
)

# Sequence analysis
seq_result = analyze_sequence("ATGCGATCG", seq_type="dna")

# Protein analysis
protein_result = analyze_protein("MVLSPADK")

# Structure analysis
structure = load_pdb_structure("protein.pdb")
contacts = find_contacts("protein.pdb", cutoff=4.0)