In-Depth Look at Affinity Chromatography Resins/Media/Purification Materials: How AI Boosts Protein Purification Efficiency

If you're following biopharmaceuticals, you've definitely heard the term 'antibody drugs.' But you might not know: from cell culture to the final product, in a typical monoclonal antibody production process, downstream purification costs often account for 50%–70% of the total production cost (higher in small-scale processes, around 30%–50% in mature large-scale processes), with Protein A affinity media being one of the single largest consumable expenses downstream.

Affinity chromatography media, affinity chromatography matrix, affinity purification media—these three terms are often used interchangeably in the industry, but if you look closely, their focus isn’t exactly the same. This article clarifies the conceptual differences between the three from a professional perspective and, combined with AI-powered innovative R&D models, explains the current state and future of this billion-dollar 'bottleneck' field.

I. Clarifying Core Concepts: Same Origin, Different Meaning

The core principle of affinity chromatography relies on the highly specific and reversible binding between biomolecules—for example, antigen-antibody, ligand-target protein, metal ion-tagged protein interactions—to precisely capture the target molecule from a complex sample system, removing other proteins and impurities to achieve efficient separation and purification. It is currently one of the most specific and effective chromatography techniques.

The three main materials we often mention—affinity chromatography media, affinity chromatography matrix, and affinity purification media—are essentially all combinations of 'matrix beads + specific ligand.' The difference mainly lies in perspective and usage:

1. Affinity chromatography media is a broad concept from a technical and material science perspective, emphasizing its complete function system as a chromatographic separation material, including the matrix material, pore structure, surface modification, ligand coupling, and all technical elements. It assesses lower-level performance indicators like hydrophilicity, physicochemical stability, and ligand coupling efficiency.

2. Affinity chromatography matrix is the common term used from an industrial application perspective, emphasizing the ready-to-use form. The media, after sieving, homogenization, and stability treatment, can be directly packed into chromatography columns and connected to purification equipment for production. It focuses more on particle uniformity, batch consistency, pressure resistance, and flow rate adaptability—all indicators directly related to production. Terms like Protein A affinity matrix, IMAC (immobilized metal affinity chromatography), and antigen-specific affinity matrix fall into this context.

3. Affinity purification media is a more functional and application-oriented term, narrower in classification, focusing on the fine purification of recombinant proteins, antibodies, functional enzymes, etc. These media are optimized for purification scenarios—resistant to acidic/alkaline cleaning, tolerant to high-salt buffers, and reusable—allowing direct processing of complex samples like cell culture fluids, serum, or fermentation supernatants, achieving high-purity target product in a single step.

In short: the matrix + ligand is the core, media is the technical term, matrix is the application form, and purification media is the functional subdivision. The three share the same origin but have different meanings, though their core structure and technical logic are consistent.

II. Market Landscape: Billion-Dollar Track, Oligopoly

How big is the market for affinity chromatography resins? Let’s look at it from a few perspectives:

From the broadest separation media market view, the global separation media market is expected to reach $12.87 billion by 2025, up 7.2% year-on-year. The biopharmaceutical field continues to be the biggest driver, with demand for monoclonal antibody and recombinant protein purification making up 55.6% of the total demand. Focusing on protein purification resins, the global market size will reach about $8.26 billion by 2025, a 12.4% increase from 2024, with monoclonal antibody purification demand accounting for 47.3% of the total market. According to QYResearch, global chromatography resin sales reached $1.951 billion in 2025 and are expected to reach $3.212 billion by 2032, with a compound annual growth rate of 7.4%.

The Chinese market is growing significantly faster than the global market. According to the Forward Industry Research Institute, the Chinese biopharmaceutical chromatography media market is expected to reach 10.8 billion RMB by 2025 and is projected to hit 13.2 billion RMB by 2026. The overall Chinese separation media market is expected to exceed 22 billion RMB in 2025, increasing its share of the global market to 17.3%. Among these, affinity chromatography resins are the largest subcategory and the most watched area for domestic substitution.

The growth drivers are very clear: the antibody drug market continues to expand. According to Sullivan's "2026 Global Antibody Drug Industry Development Blue Book," the global antibody drug market reached $254.3 billion in 2024 and is expected to exceed $675.7 billion by 2035. The more antibody drugs are sold, the more affinity resins are needed—this is an extremely certain growth logic.

However, this market has long been dominated by a few international giants. The top ten global resin suppliers account for about 71% of the market. For a long time, the core technology and pricing power of high-end Protein A resins have been controlled by foreign companies like Cytiva and Merck. Imported products are expensive, have long delivery cycles, and carry the risk of supply interruptions due to geopolitical issues.

III. Core Pain Point: The Challenge of Industrial-Scale Modification of Affinity Ligands

 Visualisation of AI-driven precision optimisation of ligand molecules

The quality of affinity resins ultimately comes down to the ligand—the affinity molecule responsible for "capturing" the target protein.

Take the most mainstream Protein A ligand as an example. It naturally exists on the surface of Staphylococcus aureus and can specifically bind to the Fc region of antibodies. But using natural Protein A directly in industrial production faces three major challenges:

First, insufficient alkali resistance. According to the alkali tolerance level of the resin, 0.1–1.0 M NaOH is usually used for in-place cleaning (CIP) to remove contaminants, deactivate viruses, and regenerate the medium. Natural Protein A is easily deactivated in strong alkaline conditions, resulting in a very short service life and higher consumable costs.

Second, high-affinity, high-specificity custom ligands are scarce. Protein A solves the purification problem for IgG antibodies, but for the growing number of new molecules—bispecific antibodies, VHH nanobodies, fusion proteins, viral vectors, membrane proteins—there often isn’t a ready-made high-affinity natural ligand. Traditional ligand screening takes a long time and is expensive, and for many niche targets, there are simply no commercial resins available.

Third, the risk of ligand leaching. Ligands are covalently attached to the resin surface, but repeated acidic and alkaline elution and CIP cleaning can cause trace amounts of ligand to leach into the product. For injectable biologics, ligand leaching is a strictly controlled safety risk. Even ppm-level leaching requires detailed safety assessment.

Traditionally, ligand optimization relies on directed evolution or rational design driven by expert experience—long cycles, high costs, and low success rates. Developing an industrial-grade ligand from discovery to mature optimization can take 2–5 years.

This is precisely where AI technology comes in.

IV. AI-Powered New Paradigm: Redefining Resin Development Logic

For the three major industry pain points mentioned above, AI protein design offers a completely new solution.

Take Shanghai Matwings Technology’s independently developed MatwingsVenus™ (Xiaowu™) platform as an example. This is an AI agent–driven conversational protein development platform. Relying on a self-built ultra-large protein database, multidimensional AI design models, and a dry-wet experimental closed-loop system, it’s shifting the development of affinity resin ligands from a "trial-and-error mode" to a "precision design mode."

1. In terms of alkali resistance modification, AI has achieved ‘precision guidance.’

How was traditional alkali resistance modification done? Random mutation libraries were created, then each variant’s activity was tested, relying on luck to screen from thousands of mutants. AI models, by learning massive protein sequence-structure-function data, can accurately predict which mutation sites are most likely to enhance the structural stability of ligands under strong alkali conditions without affecting binding activity. The search space is reduced from ‘full-scale traversal’ to ‘precise targeting,’ cutting down the experimental workload by an order of magnitude. In 2024, Matwings Technology and GenScript Pharmaceuticals collaborated to custom-develop highly alkali-resistant affinity resins using AI large models, successfully evolving single-domain antibodies into industrial-grade alkali-resistant affinity resins and achieving industrial production at a 5000-liter scale.

2. In custom ligand development, AI has greatly lowered the barriers.

For ‘non-standard’ molecules like VHH nanobodies, bispecific antibodies, and various recombinant proteins, traditional methods for developing specific affinity ligands are extremely costly and time-consuming. AI can, based on the target protein’s structural information, design or optimize highly specific ligands from scratch—without relying on animal immunization or building large libraries for screening. Some research teams combined the protein language model ESM3 with computer simulation screening to successfully develop affinity peptide ligands for lentiviral vector chromatography purification. This means the development cost of custom affinity resins will drop significantly—custom solutions that were previously only affordable for large-scale products will soon be within reach for small and medium-sized products too.

3. In controlling ligand shedding, AI optimizes coupling design at the molecular level.

Ligand shedding is caused by multiple factors, including the ligand’s intrinsic stability, coupling site, and coupling method. AI can assist in designing the optimal coupling sites and strategies—for example, selecting specific amino acids on the protein surface for site-specific coupling to ensure proper ligand orientation (improving effective binding rate) while enhancing the chemical stability of the coupling bond (reducing shedding).

Unlike traditional trial-and-error R&D, AI-driven affinity resin development creates a closed loop of ‘AI design—automated experimental validation—data feedback iteration.’ Once users define the resin’s application scenarios, performance parameters, and tolerance conditions, the MatwingsVenus™ (Xiaowu™) platform can independently complete ligand protein sequence design, structure optimization, and physicochemical property prediction. The traditional 2–5 year development cycle can be shortened to 2–6 months, with the experimental workload reduced from over a thousand mutants to just over a hundred.

V. Future Outlook: Three Certain Trends

The Development of the Affinity Chromatography Resin Industry and the Landscape of Comprehensive Purification Solutions

Looking at the affinity chromatography resin industry from the vantage point of 2026, three trends have already become very clear:

First, domestic substitution has entered deep waters.

It’s no longer about low-end substitution with just "cheap is fine" mentality, but now fully matching international mainstream products on high-end indicators like high binding capacity, stability, and long lifespan. Currently, domestic Protein A resins can achieve dynamic binding capacities of over 60 mg/mL, and after 100 cycles of CIP with 0.5M NaOH, binding capacity retention exceeds 80%, with some indicators even locally surpassing international benchmarks. As the commercialization capacity of domestic antibody drugs continues to be released, domestic resins will gradually move from the R&D and clinical stages to large-scale commercial production, and their market share will keep growing.

Second, customized demand is booming, and AI is the key factor.

Bispecifics, ADCs, VHHs, fusion proteins, enzymes, vaccines… the molecular forms of new biopharmaceuticals and bioproducts are becoming increasingly diverse, and generic Protein A resins just can’t cover all scenarios. The demand for customized affinity resins will continue to grow, but traditional methods’ high costs and long timelines limit their popularization. AI design is the key factor in reducing the cost of customization and shortening development cycles—once the cost of developing custom ligands drops low enough and the timeline short enough, the entire affinity resin market will change: moving from "a few general products dominating the market" to "customized per drug, one resin per drug".

Third, industry competition is upgrading from "selling resins" to "providing complete purification solutions."

Resins themselves are just the "hardware"; the real value lies in the service around process development, optimization, and scale-up. Future competition in resins will not only be about product performance, but also about who can help customers accelerate the process, cut costs, and scale up production faster. AI design capabilities and automated experimental platforms will become key differentiating factors for resin companies—whoever can help customers complete the full process from ligand design to process verification faster and more efficiently will hold the upper hand in next-generation competition.

Looking even further, AI-designed affinity resins could even give rise to completely new application scenarios—such as high-flow resins for continuous chromatography, low-cost customized resins for single-use processes, and special affinity resins for emerging gene therapy products, and so on. Once the technological boundary is opened, imagination is the only limit.

VI. Conclusion

Clearly defining the concepts and application differences of affinity chromatography resins, media, and purification resins is fundamental to optimizing bioprocessing and choosing the right core consumables. Affinity chromatography resins are a typical case of a 'small industry, big impact.' While the market size itself isn’t the biggest, it directly determines the production cost, product quality, and capacity ceiling of biologics — calling it one of the 'bottleneck' links in the biomanufacturing supply chain isn’t an exaggeration.

It took over a decade for domestic resins to go from 'following behind' to 'running alongside.' Now, with the entry of AI protein design, we’re seeing a new round of technological acceleration — not just catching up with international giants, but also having the chance to carve out our own path in new areas like customization and intelligent design. With the deep implementation of AI for Science, the traditional trial-and-error approach in resin development is being rewritten, and affinity chromatography resins, these 'chip-level' consumables, are set to enter an entirely new growth cycle powered by AI.