Chemical Modification Patterns Reference

Comprehensive guide to chemical modification patterns for siRNA stability, delivery, and therapeutic applications.

Overview

Chemical modifications enhance siRNA stability and reduce off-target effects while maintaining on-target potency. This guide covers built-in patterns, FDA-approved examples, and best practices for custom pattern design.

Built-In Modification Patterns

Minimal Terminal (Cost-Optimized)

File: examples/modification_patterns/minimal_terminal.json

Strategy:

  • Guide strand: 3’ terminal only (19, 20, 21)

  • Passenger strand: 5’ terminal only (1, 2)

  • Overhang: dTdT (DNA)

Properties:

  • Serum stability: ~6 hours

  • Synthesis cost: 1.1× unmodified

  • Nuclease resistance: Low-moderate

  • RISC loading: High efficiency

Best For:

  • High-throughput screening

  • Cost-sensitive experiments

  • In vitro only (cell culture with serum-free media)

  • Transient knockdown studies

Limitations:

  • Not suitable for in vivo work

  • Requires frequent media changes

  • Limited serum stability

Example:

sirnaforge design input.fasta \
  --modification-file examples/modification_patterns/minimal_terminal.json \
  --output minimal_cost.csv

Maximal Stability (Therapeutic-Grade)

File: examples/modification_patterns/maximal_stability.json

Strategy:

  • Guide strand: Complete 2’-O-methyl at all positions (1-21) + PS linkages at terminal dinucleotides

  • Passenger strand: Complete 2’-O-methyl (1-21) + 5’ PS linkages

  • Overhang: dTdT with optional PS modifications

Properties:

  • Serum stability: ~72 hours

  • Synthesis cost: 3.0× unmodified

  • Nuclease resistance: Very high

  • RISC loading: High efficiency

Best For:

  • In vivo efficacy studies

  • Therapeutic development programs

  • Preclinical toxicology

  • Long-term knockdown experiments

Requirements:

  • Specialized delivery (LNP or GalNAc conjugates)

  • HPLC purification recommended

  • Extended synthesis time (2-4 weeks)

Example:

sirnaforge workflow TTR \
  --modification-file examples/modification_patterns/maximal_stability.json \
  --genome-species human \
  --output-dir ttr_therapeutic

FDA-Approved References:

  • Patisiran (Onpattro) - TTR, 2018, hATTR amyloidosis

  • Givosiran (Givlaari) - ALAS1, 2019, Acute hepatic porphyria

FDA-Approved Example: Patisiran (Onpattro)

File: examples/modification_patterns/fda_approved_onpattro.json

First FDA-approved RNAi therapeutic (2018) targeting transthyretin (TTR) for hereditary transthyretin-mediated amyloidosis.

Guide Strand (Patisiran)

Sequence: AUGGAAUACUCUUGGUUAC

Modifications:

  • 2’-O-methyl at positions: 1, 4, 6, 11, 13, 16, 19

  • Overhang: dTdT

  • Strategic pattern optimized for stability + RISC loading

Rationale:

  • Balances nuclease resistance with guide strand activity

  • Maintains A/U-rich 5’ seed region accessibility

  • 7 modifications provide excellent stability without over-modification

Passenger Strand (Patisiran)

Sequence: GUAACCAAGAGUAUUCCAU

Modifications:

  • 2’-O-methyl at positions: 3, 8, 10, 15

  • Overhang: dTdT

  • Limited modifications to promote degradation

Rationale:

  • Fewer modifications than guide (intentional)

  • Promotes preferential RISC loading of guide strand

  • Reduces passenger strand activity/off-targets

Clinical Success

Efficacy:

  • Significant TTR reduction in Phase III trials (80% knockdown)

  • FDA approved August 2018

  • First-in-class RNAi therapeutic

Delivery:

  • Lipid nanoparticle (LNP) formulation

  • IV infusion: 0.3 mg/kg every 3 weeks

  • Hepatocyte-targeted delivery

Use This Pattern For:

  • Liver-targeted therapeutic development

  • Benchmarking modification strategies

  • Regulatory submission templates

  • Educational/training purposes

Example:

# Design TTR-targeting siRNA using Patisiran pattern
sirnaforge workflow TTR \
  --modification-file examples/modification_patterns/fda_approved_onpattro.json \
  --genome-species human \
  --output-dir ttr_patisiran_template

References: TODO: Review

  • Adams et al. (2018) NEJM 379:11-21. PMID: 30145929

  • US Patent US10060921B2

  • FDA Drug Approval Package (2018)

Modification Types Reference

2’-O-Methyl (2OMe)

Most Common Choice

  • Chemistry: Methyl group at 2’ position of ribose

  • Stability gain: ++ (good)

  • Cost factor: 1.2-1.5× per modification

  • RISC compatibility: Excellent

  • Typical positions: Alternating or custom patterns

Pros:

  • Industry standard

  • Well-characterized

  • Compatible with all synthesis platforms

  • Maintains Watson-Crick pairing

Cons:

  • Moderate cost increase

  • May require 10+ modifications for therapeutic stability

Phosphorothioate (PS)

Internucleotide Linkages

  • Chemistry: Sulfur replaces non-bridging oxygen in phosphate backbone

  • Stability gain: +++ (very good)

  • Cost factor: 1.3-1.8× per linkage

  • Typical positions: Terminal dinucleotides (5’ and 3’)

Pros:

  • Excellent nuclease resistance

  • Enhances protein binding (albumin)

  • Can improve pharmacokinetics

Cons:

  • Potential for non-specific binding

  • Synthesis complexity increases

  • Can affect duplex stability

Best Practice:

  • Limit to 2-4 linkages per strand

  • Focus on terminal positions

  • Often combined with 2’-O-methyl

2’-Fluoro (2F)

Pyrimidine-Specific

  • Chemistry: Fluorine at 2’ position (C and U only)

  • Stability gain: +++ (very good)

  • Cost factor: 1.5-2.0×

  • Typical positions: All pyrimidines

Pros:

  • Superior nuclease resistance vs 2OMe

  • Maintains duplex stability

  • Good RISC compatibility

Cons:

  • Higher synthesis cost

  • Pyrimidine-restricted (can’t modify A/G)

  • Less commonly used than 2OMe

Locked Nucleic Acid (LNA)

High-Affinity Modifications

  • Chemistry: Methylene bridge locks ribose in C3’-endo conformation

  • Stability gain: ++++ (excellent)

  • Cost factor: 2-3× per residue

  • Typical positions: Sparse (every 3-4 nt)

Pros:

  • Extremely high binding affinity

  • Superior nuclease resistance

  • Very effective at low frequency

Cons:

  • Expensive

  • Can inhibit RISC loading if over-used

  • Requires careful positioning

  • Not all synthesis providers offer

Best Practice:

  • Use sparingly (2-3 per strand maximum)

  • Avoid seed region (positions 2-8)

  • Often combined with 2’-O-methyl

2’-O-Methoxyethyl (MOE)

Alternative to 2OMe

  • Chemistry: Methoxyethyl group at 2’ position

  • Stability gain: ++ (good)

  • Cost factor: 1.5-2.0×

Pros:

  • Good nuclease resistance

  • Reduced immunogenicity vs 2OMe

  • Used in some antisense applications

Cons:

  • Less common than 2OMe for siRNA

  • Higher cost

  • Limited track record in approved therapeutics

Custom Pattern Design

Design Principles

1. Start Conservative

  • Begin with standard patterns

  • Add modifications incrementally

  • Test efficacy at each step

2. Balance Competing Goals

  • Stability ↔ Cost

  • Nuclease resistance ↔ RISC loading

  • On-target potency ↔ Off-target reduction

3. Consider Application

  • In vitro: Minimal modifications sufficient

  • In vivo (research): Standard patterns

  • Therapeutic: Maximal stability required

4. Strand Asymmetry

  • More modifications on guide strand

  • Fewer modifications on passenger strand

  • Promotes guide strand RISC loading

Pattern Template

Create custom JSON files following this structure:

{
  "pattern_name": "my_custom_pattern",
  "description": "Brief description of strategy",
  "reference": "Literature or internal study reference",

  "guide_modifications": {
    "2OMe": {
      "positions": [1, 3, 5, 7, 9],
      "strategy": "alternating",
      "rationale": "Why these positions?"
    },
    "PS": {
      "positions": [],
      "internucleotide_linkages": [[1, 2], [20, 21]],
      "rationale": "Terminal linkages for stability"
    }
  },

  "passenger_modifications": {
    "2OMe": {
      "positions": [2, 4, 6],
      "strategy": "sparse",
      "rationale": "Minimal to promote degradation"
    }
  },

  "overhang": {
    "guide_3prime": "dTdT",
    "passenger_3prime": "dTdT",
    "rationale": "Standard DNA overhangs"
  },

  "estimated_properties": {
    "serum_stability_half_life_hours": 24,
    "relative_synthesis_cost": 1.5,
    "nuclease_resistance": "moderate-high",
    "risc_loading_efficiency": "high",
    "recommended_for": "in_vitro, initial_in_vivo"
  },

  "notes": [
    "Additional context",
    "Testing recommendations",
    "Known limitations"
  ]
}

Position Selection Strategies

Alternating (Balanced)

Guide:     [1, 3, 5, 7, 9, 11, 13, 15, 17, 19]
Passenger: [2, 4, 6, 8, 10, 12, 14, 16, 18, 20]

  • Standard industry approach

  • Good stability/cost balance

  • Maintains duplex structure

Terminal-Heavy (Cost-Optimized)

Guide:     [1, 2, 19, 20, 21]
Passenger: [1, 2]

  • Minimal synthesis cost

  • Protects most vulnerable positions

  • Suitable for in vitro only

Seed-Region Preservation

Guide:     [1, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
Passenger: [full complement]

  • Avoids positions 2-8 (seed region)

  • May improve target recognition

  • Based on some miRNA-mimetic designs

Complete (Therapeutic)

Guide:     [1-21] all positions
Passenger: [1-21] all positions
+ PS linkages at terminals

  • Maximum stability

  • Therapeutic-grade

  • Highest cost

Application-Specific Recommendations

In Vitro Screening (96-well, 384-well)

Recommended Pattern: Minimal Terminal

Rationale:

  • Cost is primary concern

  • Short exposure time (<72 hours)

  • Serum-free or low-serum media

  • High-throughput compatible

Example:

sirnaforge design library.fasta \
  --modification-file examples/modification_patterns/minimal_terminal.json \
  --top-n 500 \
  --output hts_library.csv

In Vitro Validation Studies

Recommended Pattern: Standard 2’-O-Methyl

Rationale:

  • Industry standard for publications

  • Good stability in culture media

  • Reproducible results

  • Comparable to literature

Example:

sirnaforge workflow GENE \
  --modification-file examples/modification_patterns/standard_2ome.json \
  --top-n 20

In Vivo Proof-of-Concept

Recommended Pattern: Standard or Maximal Stability

Rationale:

  • Need extended half-life

  • Systemic delivery challenges

  • Nuclease-rich environment

  • Justifies higher cost

Example:

sirnaforge workflow TARGET \
  --modification-file examples/modification_patterns/maximal_stability.json \
  --genome-species mouse \
  --output-dir invivo_poc

Therapeutic Development

Recommended Pattern: Maximal Stability (Custom Optimized)

Rationale:

  • Regulatory requirements

  • Long-term efficacy needed

  • Safety/tox studies

  • Cost justified by value

Requirements:

  • GMP-grade synthesis

  • HPLC purification

  • Mass spec confirmation

  • Batch-to-batch QC

Consider:

  • GalNAc conjugation for hepatocytes

  • Lipid nanoparticle formulation

  • Antibody conjugates for targeting

  • Patent landscape review

Synthesis and Ordering

Synthesis Vendors

Major Providers:

  • Integrated DNA Technologies (IDT)

  • Thermo Fisher (Dharmacon)

  • Sigma-Aldrich

  • GenePharma

  • Biomers.net

  • TriLink BioTechnologies

Cost Estimates (2025)

Base siRNA (21bp duplex, unmodified): ~$200-400

Modifications (per strand):

  • 2’-O-methyl: +$20-50 per modification

  • Phosphorothioate: +$30-80 per linkage

  • 2’-Fluoro: +$40-100 per modification

  • LNA: +$100-200 per residue

Additional Costs:

  • HPLC purification: +$100-300

  • Mass spec QC: +$50-150

  • Bulk synthesis (5+ sequences): -30-50% discount

  • GMP-grade: 2-5× standard pricing

Pattern Cost Examples:

  • Minimal terminal: ~$250-450

  • Standard 2OMe: ~$400-600

  • Maximal stability: ~$800-1,200

  • Therapeutic + conjugate: $2,000-5,000+

Ordering Checklist

Required Information:

  1. ✅ Sequences (guide + passenger)

  2. ✅ Modification map (positions + types)

  3. ✅ Overhang specification

  4. ✅ Purification method (HPLC recommended)

  5. ✅ Scale (nmol) - typically 100-200 nmol research scale

  6. ✅ QC requirements (mass spec, HPLC traces)

Recommended Documentation:

  • Provide modification JSON file

  • Include synthesis notes from vendor

  • Archive lot numbers and QC data

  • Store at -20°C or -80°C

Timeline Expectations

  • Unmodified: 5-10 business days

  • Standard modifications: 2-3 weeks

  • Extensive modifications: 3-4 weeks

  • GMP/therapeutic grade: 6-12 weeks

  • Custom conjugates: 4-8 weeks

Best Practices

1. Design Before You Modify

✅ Do: Design unmodified siRNAs first, validate efficacy, then apply modifications

❌ Don’t: Start with heavily modified sequences without unmodified baseline

2. Test Incrementally

✅ Do: Compare minimal → standard → maximal patterns

❌ Don’t: Jump to maximal modifications without intermediate validation

3. Document Everything

✅ Do: Record modification patterns, lot numbers, QC data, storage conditions

❌ Don’t: Rely on memory or informal notes

4. Match Pattern to Application

✅ Do: Use minimal for screening, standard for validation, maximal for in vivo

❌ Don’t: Over-spend on stability you don’t need

5. Consider Delivery Method

✅ Do: Design modifications compatible with your delivery system (LNP, electroporation, etc.)

❌ Don’t: Ignore how modifications affect delivery vehicle interactions

6. Archive Modification Files

✅ Do: Version control JSON files with sequences

❌ Don’t: Recreate modification schemes from memory

Integration with siRNAforge

Command-Line Usage

# Apply pattern during design
sirnaforge design input.fasta \
  --modification-file pattern.json \
  --output designed.csv

# Apply pattern during workflow
sirnaforge workflow GENE \
  --modification-file pattern.json \
  --output-dir results

Python API

from sirnaforge.modifications import load_metadata, save_metadata_json
from sirnaforge.models.modifications import StrandMetadata, ChemicalModification

# Load built-in pattern
pattern = load_metadata("examples/modification_patterns/standard_2ome.json")

# Create custom metadata
metadata = StrandMetadata(
    id="custom_sirna_001",
    sequence="AUCGAUCGAUCGAUCGAUCGA",
    overhang="dTdT",
    chem_mods=[
        ChemicalModification(type="2OMe", positions=[1, 3, 5, 7, 9])
    ]
)

# Save for later use
save_metadata_json({"custom_sirna_001": metadata}, "my_mods.json")

Annotate Existing FASTA

# Merge modification metadata into FASTA headers
sirnaforge sequences annotate \
  candidates.fasta \
  modifications.json \
  -o candidates_annotated.fasta

Troubleshooting

Low Efficacy After Modification

Possible Causes:

  • Over-modification in seed region (positions 2-8)

  • Excessive passenger modifications reducing RISC loading

  • Delivery method incompatibility

Solutions:

  • Reduce modifications in seed region

  • Ensure guide strand preference (asymmetric modification)

  • Test unmodified sequence as control

High Cost Synthesis

Possible Causes:

  • Too many modifications

  • Exotic modification types (LNA, custom)

  • Small scale orders

Solutions:

  • Use minimal or standard patterns

  • Order multiple sequences together (bulk discount)

  • Request quotes from multiple vendors

Inconsistent Results

Possible Causes:

  • Vendor variability

  • Storage degradation

  • Batch differences

Solutions:

  • Specify HPLC purification

  • Request lot-to-lot QC data

  • Store properly (-20°C or -80°C)

  • Use consistent vendor/synthesis method

References

Key Publications

  1. Alnylam Platform Papers

    • Foster et al. (2018) “Advanced siRNA designs further improve in vivo performance” - Mol Ther 26:708-717

  2. FDA-Approved Therapeutics

    • Adams et al. (2018) “Patisiran, an RNAi therapeutic…” - NEJM 379:11-21

    • Balwani et al. (2020) “Givosiran for acute hepatic porphyria” - NEJM 382:2289-2301

  3. Modification Chemistry

    • Deleavey & Damha (2012) “Designing chemically modified oligonucleotides” - Chem Biol 19:937-954

  4. Design Principles

    • Khvorova & Watts (2017) “The chemical evolution of oligonucleotide therapies” - Nat Biotechnol 35:238-248

Regulatory Guidance

  • FDA Guidance for Industry: siRNA-Based Therapeutics (2020)

  • EMA Guideline on Quality of Oligonucleotide-Based Medicinal Products (2021)

Patents

  • US10060921B2: Patisiran (Onpattro) - Alnylam

  • US9605264B2: RNAi modifications - Alnylam

  • Multiple patents covering specific modification patterns

Contributing Patterns

To contribute new patterns to siRNAforge:

  1. Create JSON File: Follow the template structure

  2. Validate Format: Test with siRNAforge tools

  3. Document Rationale: Include references and use cases

  4. Add Examples: Provide working command-line examples

  5. Submit PR: Via GitHub with detailed description

Example Contribution:

# Test your pattern
sirnaforge design test.fasta \
  --modification-file my_new_pattern.json \
  --output test_output.csv

# Verify it works
cat test_output.csv | head -10

Document Version: 1.0 Last Updated: December 2025 Maintainer: siRNAforge Team

For questions or contributions, see GitHub repository.