A review of deep learning methods for ligand based drug virtual screening.

Drug discovery is costly and time consuming, and modern drug discovery endeavors are progressively reliant on computational methodologies, aiming to mitigate temporal and financial expenditures associated with the process. In particular, the time required for vaccine and drug discovery is prolonged during emergency situations such as the coronavirus 2019 pandemic. Recently, the performance of deep learning methods in drug virtual screening has been particularly prominent. It has become a concern for researchers how to summarize the existing deep learning in drug virtual screening, select different models for different drug screening problems, exploit the advantages of deep learning models, and further improve the capability of deep learning in drug virtual screening. This review first introduces the basic concepts of drug virtual screening, common datasets, and data representation methods. Then, large numbers of common deep learning methods for drug virtual screening are compared and analyzed. In addition, a dataset of different sizes is constructed independently to evaluate the performance of each deep learning model for the difficult problem of large-scale ligand virtual screening. Finally, the existing challenges and future directions in the field of virtual screening are presented.

A planar electronic acceptor motif contributing to NIR-II AIEgen with combined imaging and therapeutic applications.

Designing molecules with donor-acceptor-donor (D-A-D) architecture plays an important role in obtaining second near-infrared region (NIR-II, 1000-1700 nm) fluorescent dyes for biomedical applications; however, this always comes with a challenge due to very limited electronic acceptors. On the other hand, to endow NIR-II fluorescent dyes with combined therapeutic applications, trivial molecular design is indispensable. Herein, we propose a pyrazine-based planar electronic acceptor with a strong electron affinity, which can be used to develop NIR-II fluorescent dyes. By structurally attaching two classical triphenylamine electronic donors to it, a basic D-A-D module, namely Py-NIR, can be generated. The planarity of the electronic acceptor is crucial to induce a distinct NIR-II emission peaking at ∼1100 nm. The unique construction of the electronic acceptor can cause a twisted and flexible molecular conformation by the repulsive effect between the donors, which is essential to the aggregation-induced emission (AIE) property. The tuned intramolecular motions and twisted D-A pair brought by the electronic acceptor can lead to a remarkable photothermal conversion with an efficiency of 56.1% and induce a type I photosensitization with a favorable hydroxyl radical (OH˙) formation. Note that no additional measures are adopted in the molecular design, providing an ideal platform to realize NIR-II fluorescent probes with synergetic functions based on such an acceptor. Besides, the nanoparticles of Py-NIR can exhibit excellent NIR-II fluorescence imaging towards orthotopic 4T1 breast tumors in living mice with a high sensitivity and contrast. Combined with photothermal imaging and photoacoustic imaging caused by the thermal effect, the imaging-guided photoablation of tumors can be well performed. Our work has created a new opportunity to develop NIR-II fluorescent probes for accelerating biomedical applications.

Hit-to-Lead Optimization of the Natural Product Oridonin as Novel NLRP3 Inflammasome Inhibitors with Potent Anti-Inflammation Activity.

Targeting NLRP3 inflammasome with inhibitors is a novel strategy for NLRP3-driven diseases. Herein, hit compound 5 possessing an attractive skeleton was identified from our in-house database of oridonin, and then a potential lead compound 32 was obtained by optimization of 5, displaying two-digit nanomolar inhibition on NLRP3. Moreover, compound 32 showed enhanced safety index (SI) relative to oridonin (IC50 = 77.2 vs 780.4 nM, SI = 40.5 vs 8.5) and functioned through blocking ASC oligomerization and interaction of NLRP3-ASC/NEK7, thereby suppressing NLRP3 inflammasome assembly and activation. Furthermore, diverse agonists-induced activations of NLRP3 could be impeded by compound 32 without altering NLRC4 or AIM2 inflammasome. Crucially, compound 32 possessed tolerable pharmaceutical properties and significant anti-inflammatory activity in MSU-induced gouty arthritis model. Therefore, this work enriched the SAR of NLRP3 inflammasome inhibitors and provided a potential candidate for the treatment of NLRP3-associated diseases.

Modulation of Heterotypic and Homotypic Interactions to Visualize the Evolution of Organic Aggregates in a Fluorescence Turn-on Manner.

The abnormal evolution of membrane-less organelles into amyloid fibrils is a causative factor in many neurodegenerative diseases. Fundamental research on evolving organic aggregates is thus instructive for understanding the root causes of these diseases. In-situ monitoring of evolving molecular aggregates with built-in fluorescence properties is a reliable approach to reflect their subtle structural variation. To increase the sensitivity of real-time monitoring, we presented organic aggregates assembled by TPAN-2MeO, which is a triphenyl acrylonitrile derivative. TPAN-2MeO showed a morphological evolution with distinct turn-on emission. Upon rapid nanoaggregation, it formed non-emissive spherical aggregates in the kinetically metastable state. Experimental and simulation results revealed that the weak homotypic interactions between the TPAN-2MeO molecules liberated their molecular motion for efficient non-radiative decay, and the strong heterotypic interactions between TPAN-2MeO and water stabilized the molecular geometry favorable for the non-fluorescent state. After ultrasonication, the decreased heterotypic interactions and increased homotypic interactions acted synergistically to allow access to the emissive thermodynamic equilibrium state with a decent photoluminescence quantum yield (PLQY). The spherical aggregates were eventually transformed into micrometer-sized blocklike particles. Under mechanical stirring, the co-assembly of TPAN-2MeO and Pluronic F-127 formed uniform fluorescent platelets, inducing a significant enhancement in PLQY. These results decipher the stimuli-triggered structural variation of organic aggregates with concurrent sensitive fluorescence response and pave the way for a deep understanding of the evolutionary events of biogenic aggregates.

Dynamic Transition between Monomer and Excimer Phosphorescence in Organic Near-Infrared Phosphorescent Crystals.

Achieving efficient near-infrared room-temperature phosphorescence of purely organic phosphors remains scarce and challenging due to strong nonradiative decay. Additionally, the investigation of triplet excimer phosphorescence is rarely reported, despite the fact that excimer, a special emitter commonly formed in crystals with strong π-π interactions, can efficiently change the fluorescent properties of compounds. Herein, a series of dithienopyrrole derivatives with low triplet energy levels and stable triplet states, exhibiting persistent near-infrared room-temperature phosphorescence, is developed. Via the modification of halogen atoms, the crystals display tunable emissions of monomers from 645 to 702 nm, with a maximum lifetime of 3.68 ms under ambient conditions. Notably, excimer phosphorescence can be switched on at low temperatures, enabled by noncovalent interactions rigidifying the matrix and stabilizing triplet excimer. Unprecedentedly, the dynamic transition process is captured between the monomer and excimer phosphorescence with temperature variations, revealing that the unstable triplet excimers in crystals with a tendency to dissociate can result in the effective quench of room-temperature phosphorescence. Excited state transitions across varying environments are elucidated, interpreting the structural dynamics of the triplet excimer and demonstrating strategies for devising novel near-infrared phosphors.

MFD-GDrug: multimodal feature fusion-based deep learning for GPCR-drug interaction prediction.

The accurate identification of drug-protein interactions (DPIs) is crucial in drug development, especially concerning G protein-coupled receptors (GPCRs), which are vital targets in drug discovery. However, experimental validation of GPCR-drug pairings is costly, prompting the need for accurate predictive methods. To address this, we propose MFD-GDrug, a multimodal deep learning model. Leveraging the ESM pretrained model, we extract protein features and employ a CNN for protein feature representation. For drugs, we integrated multimodal features of drug molecular structures, including three-dimensional features derived from Mol2vec and the topological information of drug graph structures extracted through Graph Convolutional Neural Networks (GCN). By combining structural characterizations and pretrained embeddings, our model effectively captures GPCR-drug interactions. Our tests on leading GPCR-drug interaction datasets show that MFD-GDrug outperforms other methods, demonstrating superior predictive accuracy.

AttentionMGT-DTA: A multi-modal drug-target affinity prediction using graph transformer and attention mechanism.

The accurate prediction of drug-target affinity (DTA) is a crucial step in drug discovery and design. Traditional experiments are very expensive and time-consuming. Recently, deep learning methods have achieved notable performance improvements in DTA prediction. However, one challenge for deep learning-based models is appropriate and accurate representations of drugs and targets, especially the lack of effective exploration of target representations. Another challenge is how to comprehensively capture the interaction information between different instances, which is also important for predicting DTA. In this study, we propose AttentionMGT-DTA, a multi-modal attention-based model for DTA prediction. AttentionMGT-DTA represents drugs and targets by a molecular graph and binding pocket graph, respectively. Two attention mechanisms are adopted to integrate and interact information between different protein modalities and drug-target pairs. The experimental results showed that our proposed model outperformed state-of-the-art baselines on two benchmark datasets. In addition, AttentionMGT-DTA also had high interpretability by modeling the interaction strength between drug atoms and protein residues. Our code is available at https://github.com/JK-Liu7/AttentionMGT-DTA.

Effects of Different Motion Parameters on the Interaction of Fish School Subsystems.

For a long time, fish school swimming has attracted a great deal of attention in biological systems, as fish schools can have complex hydrodynamic effects on individuals. This work adopted a non-iterative, immersed boundary-lattice Boltzmann method (IB-LBM). A numerical simulation of two-dimensional three-degree-of-freedom self-propelled fish, in side-by-side, staggered, and triangle formations, was conducted by adjusting spacing and motion parameters. A comprehensive analysis of individual speed gains and energy efficiencies in these formations was carried out. Furthermore, an analysis of the hydrodynamic characteristics of fish schools was performed, using instantaneous vorticity profiles and pressure fields. Certain studies have shown that passive interactions between individuals cannot always bring hydrodynamic benefits. The swimming efficiency of side-by-side formations in the same phase gradually increases as the distance decreases, but it also brings certain burdens to individuals when the phases are different. This paper also shows that the roles of passive interactions, spacing, and deflections affect fish school subsystems differently. When the low-pressure areas created by a wake vortex act on one side of an individual's body, the tail-end fish are good at gaining hydrodynamic benefits from it. This effect is not universal, and the degree to which individuals benefit from changes in exercise parameters varies. This study provides a theoretical basis for bioinspired robots, as well as providing certain insights into the mechanism of collective biological movement.

Integrating Anion-π+ Interaction and Crowded Conformation to Develop Multifunctional NIR AIEgen for Effective Tumor Theranostics via Hippo-YAP Pathway.

The technology of aggregation-induced emission (AIE) presents a promising avenue for fluorescence imaging-guided photodynamic cancer therapy. However, existing near-infrared AIE photosensitizers (PSs) frequently encounter limitations, including tedious synthesis, poor tumor retention, and a limited understanding of the underlying molecular biology mechanism. Herein, an effective molecular design paradigm of anion-π+ interaction combined with the inherently crowded conformation that could enhance fluorescence efficacy and reactive oxygen species generation was proposed through a concise synthetic method. Mechanistically, upon photosensitization, the Hippo signaling pathway contributes to the death of melanoma cells and promotes the nuclear location of its downstream factor, yes-associated protein, which regulates the transcription and expression of apoptosis-related genes. The finding in this study would trigger the development of high-performance and versatile AIE PSs for precision cancer therapy based on a definite regulatory mechanism.

Type I Photosensitization with Strong Hydroxyl Radical Generation in NIR Dye Boosted by Vigorous Intramolecular Motions for Synergistic Therapy.

Development of type I photosensitizers (PSs) with strong hydroxyl radical (· OH) formation is particularly important in the anaerobic tumor treatment. On the other hand, it is challenging to obtain an efficient solid-state intramolecular motion to promote the development of molecular machine and molecular motor. However, the relationship between them is never revealed. In this work, a pyrazine-based near-infrared type I PS with remarkable donor-acceptor effect is developed. Notably, the intramolecular motions are almost maximized by the combination of intramolecular and intermolecular engineering to simultaneously introduce the unlimited bond stretching vibration and boost the group rotation. The photothermal conversion caused by the intramolecular motions is realized with efficiency as high as 86.8%. The D-A conformation of PS can also induce a very small singlet-triplet splitting of 0.07 eV, which is crucial to promote the intersystem crossing for the triplet sensitization. Interestingly, its photosensitization is closely related to the intramolecular motions, and a vigorous motion may give rise to a strong · OH generation. In view of its excellent photosensitization and photothermal behavior, the biocompatible PS exhibits a superior imaging-guided cancer synergistic therapy. This work stimulates the development of advanced PS for the biomedical application and solid-state intramolecular motions.

Substrate-Responsive Pillar[5]arene-Based Organic Room-Temperature Phosphorescence.

Room-temperature phosphorescence (RTP) is a photophysical phenomenon typically associated with a long-lived emission that can be detected by the unaided eye. Several natural proteins display RTP, as do certain artificial polymers. In both cases, the RTP is ascribed to effective intramolecular through-space electronic communication. However, small molecules with internal electronic communication that enable RTP are relatively rare. Herein, we describe an alkyl halide-responsive RTP system consisting of a meta-formylphenyl-bearing pillar[5]arene derivative that supports effective through-space charge transfer (TSCT) within the pillararene cavity. Treatment with bromoethane, a heavy atom-containing guest for the pillar[5]arene host, serves to enhance the emission. An isomeric para-formylphenyl-bearing pillar[5]arene system proved ineffective in producing an RTP effect. Quantum chemical calculations based on single-crystal X-ray diffraction analyses provided insights into the structural determinants governing TSCT between the 1,4-dimethoxybenzene donor units and the formylphenyl groups of the pillar[5]arene, as well as the associated energy gaps and intersystem crossing channels. We believe that the present system and the associated mechanistic analysis provide the foundation for design of new small molecule with tunable RTP features.

Multifunctional Fluorescent Main-Chain Charged Polyelectrolytes Synthesized by Cascade C-H Activation/Annulation Polymerizations.

Fluorescent polyelectrolytes have attracted enormous attention as functional polymer materials. In contrast with the widely studied conjugated polyelectrolytes with ionic groups in side chains, fluorescent main-chain charged polyelectrolytes (MCCPs) have rarely been explored due to the large synthetic difficulty. Herein, we develop a facile and atom-economical N-heterocyclic carbene-directed cascade C-H activation/annulation polymerization strategy that can transform readily available imidazolium substrates and internal diynes into multifunctional fluorescent MCCPs with complex structures and high molecular weights (absolute Mn up to 135 600) in nearly quantitative yields. The presence of multisubstituted polycyclic N-heteroaromatic cations in polymer backbones endow the obtained MCCPs with excellent solution processability, high thermal stability, and dual-state efficient fluorescence in both solution and aggregate states. Benefiting from the strong electron-withdrawing capability of the cationic heterocycles in main chains, multicolored aggregate-state fluorescence can be readily achieved by modifying the substituents around the cationic ring-fused core. Taking advantage of the good photosensitivity of the fluorescent MCCP thin films, multiscale and high-resolution fluorescent photopatterns with different colors can be facilely prepared with potential applications in optical display devices and anticounterfeiting systems. Moreover, the strong electrostatic interactions of these cationic MCCPs with anionic polyelectrolytes enable them to form multicolored fluorescent interfacial polyelectrolyte complexation microfibers with directly visualized internal structures. Such flexible microfibers can be further made into diversified forms of fiber-based macroscopic patterns or painting.

Hierarchical Supramolecular Self-Assembly: Fabrication and Visualization of Multiblock Microstructures.

Due to the fast dynamics and re-equilibration of supramolecular self-assembly, bottom-up molecular strategies to fabricate well-defined and controllable multiblock structures are rare. Herein, we propose a new concept for fabrication of fluorescent multiblock microcolumns containing 1 to 7 blocks via hierarchical supramolecular self-assembly based on cucurbit[8]uril (CB[8]), NaBr and an AIEgen guest. Through the complexation between CB[8] and different numbers of AIEgen guests (2, 1, 0), the competitive displacement caused by the binding of the sodium cation to the CB[8] portal, and the reversible assembly of positively charged guests in salt solutions, one-pot hierarchical supramolecular self-assembly is realized. The molecular structure of each block is analyzed by single-crystal X-ray diffraction. The AIEgen enables the self-assembly of multiblocks to be visualized, understood, and regulated.

DNA-binding protein prediction based on deep transfer learning.

The study of DNA binding proteins (DBPs) is of great importance in the biomedical field and plays a key role in this field. At present, many researchers are working on the prediction and detection of DBPs. Traditional DBP prediction mainly uses machine learning methods. Although these methods can obtain relatively high pre-diction accuracy, they consume large quantities of human effort and material resources. Transfer learning has certain advantages in dealing with such prediction problems. Therefore, in the present study, two features were extracted from a protein sequence, a transfer learning method was used, and two classical transfer learning algorithms were compared to transfer samples and construct data sets. In the final step, DBPs are detected by building a deep learning neural network model in a way that uses attention mechanisms.

Through-Space Interaction of Tetraphenylethylene: What, Where, and How.

Electronic conjugation through covalent bonds is generally considered as the basis for the electronic transition of organic luminescent materials. Tetraphenylethylene (TPE), an efficient fluorophore with aggregation-induced emission character, fluoresces blue emission in the aggregate state, and such photoluminescence is always ascribed to the through-bond conjugation (TBC) among the four phenyl rings and the central C═C bond. However, in this work, systematic spectroscopic studies and DFT theoretical simulation reveal that the intramolecular through-space interaction (TSI) between two vicinal phenyl rings generates the bright blue emission in TPE but not the TBC effect. Furthermore, the evaluation of excited-state decay dynamics suggests the significance of photoinduced isomerization in the nonradiative decay of TPE in the solution state. More importantly, different from the traditional qualitative description for TSI, the quantitative elucidation of the TSI is realized through the atoms-in-molecules analysis; meanwhile, a theoretical solid-state model for TPE and other multirotor systems for studying the electronic configuration is preliminarily established. The mechanistic model of TSI delineated in this work provides a new strategy to design luminescent materials beyond the traditional theory of TBC and expands the quantum understanding of molecular behavior to the aggregate level.

Oxygen Quenching-Resistant Nanoaggregates with Aggregation-Induced Delayed Fluorescence for Time-Resolved Mapping of Intracellular Microviscosity.

Microviscosity is a fundamental parameter in the biophysics of life science and governs numerous cellular processes. Thus, the development of real-time quantitative monitoring of microviscosity inside cells is important. The traditional probes for detecting microviscosity via time-resolved luminescence imaging (TRLI) are generally disturbed by autofluorescence or surrounding oxygen in cells. Herein, we developed loose packing nanoaggregates with aggregation-induced delayed fluorescence (FKP-POA and FKP-PTA) and free from the effect of oxygen and autofluorescence for viscosity mapping via TRLI. The feasibility of FKP-PTA nanoparticles (NPs) for microviscosity mapping through TRLI was demonstrated by monitoring the variation of microviscosity inside HepG2 cancer cells, which demonstrated a value change from 14.9 cP to 216.9 cP during the apoptosis. This indicates that FKP-PTA NP can be used as a probe for cellular microviscosity mapping to help people to understand the physiologically dynamic microenvironment. The present results are expected to promote the advancement of diagnostic and therapeutic methods to cope with related diseases.

Click Synthesis Enabled Sulfur Atom Strategy for Polymerization-Enhanced and Two-Photon Photosensitization.

Facile tailoring of photosensitizers (PSs) with advanced and synergetic properties is highly expected to broaden and deepen photodynamic therapy (PDT) applications. Herein, a catalyst-free thiol-yne click reaction was employed to develop the sulfur atom-based PSs by using the in situ formed sulfur "heavy atom effect" to enhance the intersystem crossing (ISC), while such an effect can be remarkably magnified by the polymerization. The introduction of a tetraphenylpyrazine-based aggregation-induced emission (AIE) unit was also advantageous in PS design by suppressing their non-radiative decay to facilitate the ISC in the aggregated state. Besides, the resulting sulfur atom electron donor, together with a double-bond π bridge and AIE electron acceptor, created a donor-π-acceptor (D-π-A) molecular system with good two-photon excitation properties. Combined with the high singlet oxygen generation efficiency, the fabricated polymer nanoparticles exhibited an excellent in vitro two-photon-excited PDT towards cancer cells, therefore possessing a huge potential for the deep-tissue disease therapy.

A platform method for plasmid isoforms analysis by capillary gel electrophoresis.

In the production of novel biological products, plasmids are often engineered into delivery vectors for target genes, which can be used directly as vaccines or as intermediate products for gene/cell therapy. Plasmid DNA exists in several topological forms such as supercoiled, linear, and open circular. As supercoiled plasmid shows the highest efficiency in transfecting eukaryotic cells, the content of supercoiled plasmids becomes an important indicator of plasmid quality. CGE is an effective analysis method for separating different topological structures of plasmids. For the purpose of providing plasmid manufacturers and regulatory agencies with an efficient and readily used tool for monitoring the quality of plasmids, this article identifies the optimal separation and detection conditions of CGE, presents a platform-based plasmid analytical method, and uses plasmid of different sizes to verify the feasibility of this method. In terms of detector, the LIF detector has obvious advantages over the ultraviolet detector in sensitivity and resolution. Using the optimal CE condition (10× gel buffer), baseline separation of different topological forms and impurities can be achieved for different plasmid sizes (5.9, 7.8, 15.4 kb). In addition, 6.5 kb plasmid was used to compare the different separation technologies such as CGE-LIF, ion exchange chromatography and agarose gel electrophoresis. The result shows that CGE-LIF can provide better resolution and quantitation accuracy than ion exchange chromatography and agarose gel electrophoresis. CGE-LIF, as a quick and convenient method to separate and quantify plasmids, has the advantages of high sensitivity, high resolution, and high quantitative accuracy. Therefore, it is ideal for analysis of plasmids with different sizes, and it can also be used as a platform method for manufacturers and regulatory agencies to monitor the purity and stability of plasmids.

Identification of new potent anticancer derivatives through simplifying the core structure and modification on their 14- hydroxyl group from oridonin.

The natural product oridonin has the potential to be a broad-spectrum antineoplastic agent. To develop oridonin analogues with high potency, a series of novel oridonin analogues were designed and synthesized by removing the multiple hydroxyl groups of parent compound. The representative analogues 14, 19, and 26 exhibited potent anticancer effects against K562, MDA-MB-231, SMMC-7721, and MCF-7 cells. Further structural modification on their 14-OH generated more potent derivatives 16n, 21d, and 28d respectively, in which the IC50 value of compound 16n was 50-fold more potent than parent oridonin in K562 cells. Furthermore, compound 16n significantly induced the cell cycle arrest of K562 cells at the G2 phase and increased the fraction of apoptotic cells. Importantly, compounds 16n, 21d, and 28d exhibited good antitumor activities in H22 allograft mice in vivo. These results suggest that compounds 16n, 21d, and 28d deserve further development as promising candidates for the treatment of cancers.

[Quantitative analysis of nine types of virus-like particles in human papilloma virus bulk by size-exclusion chromatography].

Cervical cancer is the fourth most common cancer among women. Human papilloma virus (HPV) is the most common cause of cervical cancer which accounts for 5% of all human cancers and results in about 528000 cases and 266000 deaths every year. HPV vaccines are considered the most effective strategy for the prevention of HPV infection and cervical carcinoma. Since 2006, three prophylactic vaccines against HPV have been available on the market, including bivalent vaccines, quadrivalent vaccines, and nine-valent vaccines. Among them, nine-valent vaccines have been reported to be the most effective. They can prevent 97% of the high-grade pre-cancer lesions. Virus-like particles (VLPs), which are arranged as 360 copies of capsid proteins L1, are the only antigens of the HPV vaccine. Nine-valent HPV vaccines are prepared by mixing nine types of VLPs with adjuvants. Thus, the quality of the VLPs, including their stability and content in the HPV bulk, is very important for developing HPV vaccines. In this study, a method was developed for the determination of the nine types of VLPs (HPV6/11/16/18/31/33/45/52/58) in HPV bulk by size exclusion chromatography (SEC). The parameters of this method were optimized in terms of column brand, pore size of stationary phase particles, buffer concentration, and pH value. SHIMSEN Ankylo SEC-300 column (300 mm×7.8 mm, 3 µm) combined with a buffer aqueous solution containing 300 mmol/L NaCl and 50 mmol/L phosphate (pH 7.0) was utilized to separate the VLPs from the matrix since a narrow peak shape and good repeatability for VLPs could be obtained with this column and mobile phase. The optimized method had a wide linear range, good repeatability (RSDs of peak area were not more than 5.0%), and a satisfactory sensitivity (LOQs in the range of 4.58-15.24 µg/mL). The optimized method was used to determine the VLPs in the HPV bulk. The LOQs of the current method were much lower than the content of the nine types of VLPs in the HPV bulk, indicating that this method was sensitive enough for the determination of the nine types of VLPs in the HPV bulk. The method was also used to determine the VLPs in an HPV bulk that had been stored at 4 ℃ for one week. A decrease in the nine types of VLPs in the range of 10.0%-62.6% was observed after they were stored at 4 ℃ for one week. An HPV vaccine was prepared by mixing the VLPs with an adjuvant. Thereafter, the VLPs were adsorbed on the surface of the adjuvant. The developed method was applied to determine the free VLPs in twelve batches of HPV vaccines from three different manufacturers. No obvious free protein was detected in the twelve batches of the HPV vaccines from the three manufacturers, indicating that VLPs from these manufactures react well with their aluminum adjuvant. Folin-phenol (Lowry assay) is commonly used for the determination of proteins in vaccines. It is based on the reduction of phosphomolybdotungstic mixed acid chromogen in the phosphomolybdotungstic reagent, which results in an absorbance maximum at 650 nm. The Lowry method was sensitive to interfering substances. Most interfering substances caused a lower color yield, while some detergents caused a slight increase in color. To reduce the effect of the interfering substances, a procedure for precipitating the proteins was usually required before the sample was tested. Thus, the Lowry assay is complex, time-consuming, and of low selectivity. Compared to the Lowry method, the method we developed is simpler and more automatic. It is a high-throughput method of determining VLPs. It can be used to determine VLPs in HPV bulk and free VLPs in HPV vaccines.