Learning self-supervised molecular representations for drug-drug interaction prediction.

Drug-drug interactions (DDI) are a critical concern in healthcare due to their potential to cause adverse effects and compromise patient safety. Supervised machine learning models for DDI prediction need to be optimized to learn abstract, transferable features, and generalize to larger chemical spaces, primarily due to the scarcity of high-quality labeled DDI data. Inspired by recent advances in computer vision, we present SMR-DDI, a self-supervised framework that leverages contrastive learning to embed drugs into a scaffold-based feature space. Molecular scaffolds represent the core structural motifs that drive pharmacological activities, making them valuable for learning informative representations. Specifically, we pre-trained SMR-DDI on a large-scale unlabeled molecular dataset. We generated augmented views for each molecule via SMILES enumeration and optimized the embedding process through contrastive loss minimization between views. This enables the model to capture relevant and robust molecular features while reducing noise. We then transfer the learned representations for the downstream prediction of DDI. Experiments show that the new feature space has comparable expressivity to state-of-the-art molecular representations and achieved competitive DDI prediction results while training on less data. Additional investigations also revealed that pre-training on more extensive and diverse unlabeled molecular datasets improved the model's capability to embed molecules more effectively. Our results highlight contrastive learning as a promising approach for DDI prediction that can identify potentially hazardous drug combinations using only structural information.

Mating Assay: Plating Below a Cell Density Threshold is Required for Unbiased Estimation of Plasmid Conjugation Frequency of RP4 Transfer Between E. coli Strains.

Mating assays are common laboratory experiments for measuring the conjugation frequency, i.e. efficiency at which a plasmid transfers from a population of donor cells to a population of recipient cells. Selective plating remains a widely used quantification method to enumerate transconjugants at the end of such assays. However, conjugation frequencies may be inaccurately estimated because plasmid transfer can occur on transconjugant-selective plates rather than only during the intended mating duration. We investigated the influence of cell density on this phenomenon. We conducted mating experiments with IncPα plasmid RP4 harbored in Escherichia coli at a fixed cell density and mating conditions, inoculated a serial dilution of the mating mixture on transconjugant-selective plates or in transconjugant-selective broth, and compared the results to a model of cell-to-cell distance distribution. Our findings suggest that irrespective of the mating mode (liquid vs solid), the enumeration of transconjugants becomes significantly biased if the plated cell density exceeds 28 Colony Forming Unit (CFU)/mm2 (or 1.68•105 CFU/standard 9 cm Petri dish). This threshold is determined with a 95% confidence interval of ± 4 CFU/mm2 (± 2.46•104 CFU/standard 9 cm Petri dish). Liquid mating assays were more sensitive to this bias because the conjugation frequency of RP4 is several orders of magnitude lower in suspension compared to surface mating. Therefore, if selective plating is used, we recommend to plate at this density threshold and that negative controls are performed where donors and recipients are briefly mixed before plating at the same dilutions as for the actual mating assay. As an alternative, a liquid enumeration method can be utilized to increase the signal-to-noise ratio and allow for more accurate enumeration of transconjugants.

Rapid detection of viable microbes with 5-cyano-2,3-di-(p-tolyl)tetrazolium chloride and 5(6)-carboxyfluorescein diacetate using a fibre fluorescence spectroscopy system.

AIMS: To assess the efficacy of two commercially available viability dyes, 5-cyano-2,3-di-(p-tolyl)tetrazolium chloride (CTC) and 5(6)-carboxyfluorescein diacetate (CFDA), in reporting on viable cell concentration and species using an all-fibre fluorometer. METHODS AND RESULTS: Four bacterial species (two Gram-positive and two Gram-negative) commonly associated with food poisoning or food spoilage (Escherichia coli, Salmonella enterica, Staphylococcus aureus, and Bacillus cereus) were stained with CTC or CFDA and the fibre fluorometer was used to collect full fluorescence emission spectra. A good correlation between concentration and fluorescence intensity was found for Gram-negative bacteria between 107 and 108 colony-forming units (CFU) ml-1. There was no correlation with concentration for Gram-positive bacteria; however, the information in the CTC and CFDA spectra shows the potential to distinguish Gram-negative cells from Gram-positive cells, although it may simply reflect the overall bacterial metabolic activity under staining conditions from this study. CONCLUSIONS: The limit of detection (LoD) is too high in the dip-probe approach for analysis; however, the development of an approach measuring the fluorescence of single cells may improve this limitation. The development of new bacteria-specific fluorogenic dyes may also address this limitation. The ability to differentiate bacteria using these dyes may add value to measurements made to enumerate bacteria using CTC and CFDA.

Counting Phylogenetic Networks with Few Reticulation Vertices: Galled and Reticulation-Visible Networks.

We give exact and asymptotic counting results for the number of galled networks and reticulation-visible networks with few reticulation vertices. Our results are obtained with the component graph method, which was introduced by L. Zhang and his coauthors, and generating function techniques. For galled networks, we in addition use analytic combinatorics. Moreover, in an appendix, we consider maximally reticulated reticulation-visible networks and derive their number, too.

Fragment-based discovery of new potential DNMT1 inhibitors integrating multiple pharmacophore modeling, 3D-QSAR, virtual screening, molecular docking, ADME, and molecular dynamics simulation approaches.

DNA methyl transferases (DNMTs) are one of the crucial epigenetic modulators associated with a wide variety of cancer conditions. Among the DNMT isoforms, DNMT1 is correlated with bladder, pancreatic, and breast cancer, as well as acute myeloid leukemia and esophagus squamous cell carcinoma. Therefore, the inhibition of DNMT1 could be an attractive target for combating cancers and other metabolic disorders. The disadvantages of the existing nucleoside and non-nucleoside DNMT1 inhibitors are the main motive for the discovery of novel promising inhibitors. Here, pharmacophore modeling, 3D-QSAR, and e-pharmacophore modeling of DNMT1 inhibitors were performed for the large fragment database screening. The resulting fragments with high dock scores were combined into molecules. The current study revealed several constitutional pharmacophoric features that can be essential for selective DNMT1 inhibition. The fragment docking and virtual screening identified 10 final hit molecules that exhibited good binding affinities in terms of docking score, binding free energies, and acceptable ADME properties. Also, the modified lead molecules (GL1b and GL2b) designed in this study showed effective binding with DNMT1 confirmed by their docking scores, binding free energies, 3D-QSAR predicted activities and acceptable drug-like properties. The MD simulation studies also suggested that leads (GL1b and GL2b) formed stable complexes with DNMT1. Therefore, the findings of this study can provide effective information for the development/identification of novel DNMT1 inhibitors as effective anticancer agents.

YOLO-V5 based deep learning approach for tooth detection and segmentation on pediatric panoramic radiographs in mixed dentition.

OBJECTIVES: In the interpretation of panoramic radiographs (PRs), the identification and numbering of teeth is an important part of the correct diagnosis. This study evaluates the effectiveness of YOLO-v5 in the automatic detection, segmentation, and numbering of deciduous and permanent teeth in mixed dentition pediatric patients based on PRs. METHODS: A total of 3854 mixed pediatric patients PRs were labelled for deciduous and permanent teeth using the CranioCatch labeling program. The dataset was divided into three subsets: training (n = 3093, 80% of the total), validation (n = 387, 10% of the total) and test (n = 385, 10% of the total). An artificial intelligence (AI) algorithm using YOLO-v5 models were developed. RESULTS: The sensitivity, precision, F-1 score, and mean average precision-0.5 (mAP-0.5) values were 0.99, 0.99, 0.99, and 0.98 respectively, to teeth detection. The sensitivity, precision, F-1 score, and mAP-0.5 values were 0.98, 0.98, 0.98, and 0.98, respectively, to teeth segmentation. CONCLUSIONS: YOLO-v5 based models can have the potential to detect and enable the accurate segmentation of deciduous and permanent teeth using PRs of pediatric patients with mixed dentition.

New Genetic Determinants for qPCR Identification and the Enumeration of Selected Lactic Acid Bacteria in Raw-Milk Cheese.

Lactic acid bacteria (LAB) play an important role in the ripening of cheeses and contribute to the development of the desired profile of aroma and flavor compounds. Therefore, it is very important to monitor the dynamics of bacterial proliferation in order to obtain an accurate and reliable number of their cells at each stage of cheese ripening. This work aimed to identify and conduct a quantitative assessment of the selected species of autochthonous lactic acid bacteria from raw cow's milk cheese by the development of primers and probe pairs based on the uniqueness of the genetic determinants with which the target microorganisms can be identified. For that purpose, we applied real-time quantitative PCR (qPCR) protocols to quantify Lactobacillus delbrueckii subsp. bulgaricus, Streptococcus thermophilus, and Lactococcus lactis subsp. cremoris cells in cheese directly after production and over three-month and six-month ripening periods. While L. lactis subsp. cremoris shows good acidification ability and the ability to produce antimicrobial compounds, L. delbrueckii subsp. bulgaricus has good proteolytic ability and produces exo-polysaccharides, and S. thermophilus takes part in the formation of the diacetyl flavor compound by metabolizing citrate to develop aroma, they all play an important role in the cheese ripening. The proposed qPCR protocols are very sensitive and reliable methods for a precise enumeration of L. delbrueckii subsp. bulgaricus, S. thermophilus, and L. lactis subsp. cremoris in cheese samples.

The crucial importance of accurate enumeration to bringing public health nurses out from behind the "Other" category.

Public health nursing is a unique field of nursing with specialized skills, roles, and functions designed to address disease prevention and health promotion of populations and to respond to emerging health crisis such the COVID-19 pandemic. Despite the unique role and professional responsibilities of public health nurses, they are not identified as a distinct nursing specially by the U.S. Bureau of Labor Statistics workforce data. This is problematic as accurate enumeration of public health nursing can supply the necessary data to identify gaps of these essential professionals. To effectively address this gap and have the capacity to identify public health nursing workforce needs, a method to define, describe, and enumerate the public health nursing workforce nationally with a unique split Standard Occupational Classification is necessary. Further, the Centers for Disease Control and Prevention's Public Health and Data Authority must have the ability to coordinate data reporting on the public health workforce, support standardization, and streamline annual enumeration.

The State of the US Public Health Workforce: Ongoing Challenges and Future Directions.

Between the 2009 Great Recession and the onset of the COVID-19 pandemic, the US state and local governmental public health workforce lost 40,000 jobs. Tens of thousands of workers also left during the pandemic and continue to leave. As governmental health departments are now receiving multimillion-dollar, temporary federal investments to replenish their workforce, this review synthesizes the evidence regarding major challenges that preceded the pandemic and remain now. These include the lack of the field's ability to readily enumerate and define the governmental public health workforce as well as challenges with the recruitment and retention of public health workers. This review finds that many workforce-related challenges identified more than 20 years ago persist in the field today. Thus, it is critical that we look back to be able to then move forward to successfully rebuild the workforce and assure adequate capacity to protect the public's health and respond to public health emergencies.

Estimating the Primary Care Workforce for Medicare Beneficiaries Using an Activity-Based Approach.

BACKGROUND: The enumeration of the primary care workforce relies on potentially inaccurate specialty designations sourced from licensure registries and clinician surveys. OBJECTIVE: To use an activity-based measure of primary care to estimate the number of physicians, nurse practitioners (NPs), and physician assistants (PAs) providing primary care to Medicare beneficiaries. DESIGN: Observational study using Medicare fee-for-service (FFS) claims data. SUBJECTS: All clinicians in the US billing Medicare in 2019 and their fee-for-service Medicare patients. MAIN MEASURES: We construct three measures that together distinguish primary care from specialty clinicians: (1) presence of evaluation and management (E&M) services in a setting consistent with primary care, (2) the dispersion of clinical care across International Classification of Diseases-10 (ICD-10) chapters, and (3) the extent of provided services that are atypical of primary care (e.g., surgical procedure). We apply parameters to the measures to identify the clinicians likely providing primary care and compare the resulting classifications across provider type. KEY RESULTS: Of physicians with at least 50 Medicare beneficiaries, 19-22% provide primary care. Of medical generalists (i.e., family medicine, internal medicine) with at least 50 beneficiaries, 61-68% provide primary care. We estimate that 40-45% of NPs and 27-30% of PAs meeting the panel size threshold are primary care providers in FFS Medicare. CONCLUSIONS: Our findings suggest that based on a primary care practice style, the number of primary care physicians in FFS Medicare is likely smaller than conventional estimates. However, compared to prior estimates, the number of primary care NPs is larger and the number of PAs is similar.

Ecological factors shape quantitative decision-making in coyotes.

Much research has focused on the development and evolution of cognition in the realm of numerical knowledge in human and nonhuman animals but often fails to take into account ecological realities that, over time, may influence and constrain cognitive abilities in real-life decision-making. Cognitive abilities such as enumerating and timing are central to many psychological and ecological models of behavior, yet our knowledge of how these are affected by environmental fluctuations remains incomplete. Our research bridges the gap between basic cognitive research and ecological decision-making. We used coyotes (Canis latrans) as a model animal system to study decision-making about smaller, more proximal food rewards and larger, more distant food rewards; we tested animals across their four reproductive cycle phases to examine effects of ecological factors such as breeding status and environmental risk on quantitative performance. Results show that coyotes, similar to other species, spatially discount food rewards while foraging. The degree to which coyotes were sensitive to the risk of obtaining the larger food reward, however, depended on the season in which they completed the foraging task, the presence of unfamiliar humans (i.e., risk), and the presence of conspecifics. Importantly, our results support that seasonal variations drive many differences in nonhuman animal behavior and cognition (e.g., hibernation, breeding, food resource availability). Further, it may be useful in the future to extend this work to humans because seasons may influence human cognition as well, and this remains unexplored in the realms of enumeration, timing, and spatial thinking.

Exploration of model misspecification in latent class methods for longitudinal data: Correlation structure matters.

Modeling longitudinal trajectories and identifying latent classes of trajectories is of great interest in biomedical research, and software to identify latent classes of such is readily available for latent class trajectory analysis (LCTA), growth mixture modeling (GMM) and covariance pattern mixture models (CPMM). In biomedical applications, the level of within-person correlation is often non-negligible, which can impact the model choice and interpretation. LCTA does not incorporate this correlation. GMM does so through random effects, while CPMM specifies a model for within-class marginal covariance matrix. Previous work has investigated the impact of constraining covariance structures, both within and across classes, in GMMs-an approach often used to solve convergence problems. Using simulation, we focused specifically on how misspecification of the temporal correlation structure and strength, but correct variances, impacts class enumeration and parameter estimation under LCTA and CPMM. We found (1) even in the presence of weak correlation, LCTA often does not reproduce original classes, (2) CPMM performs well in class enumeration when the correct correlation structure is selected, and (3) regardless of misspecification of the correlation structure, both LCTA and CPMM give unbiased estimates of the class trajectory parameters when the within-individual correlation is weak and the number of classes is correctly specified. However, the bias increases markedly when the correlation is moderate for LCTA and when the incorrect correlation structure is used for CPMM. This work highlights the importance of correlation alone in obtaining appropriate model interpretations and provides insight into model choice.

Application of MALDI-TOF MS for enumerating bacterial constituents of defined consortia.

Characterization of live biotherapeutic product (LBP) batches typically includes a measurement of viability, such as colony forming units (CFU). However, strain-specific CFU enumeration assays can be complicated by the presence of multiple organisms in a single product with similar growth requirements. To overcome specific challenges associated with obtaining strain-specific CFU values from multi-strain mixtures, we developed a method combining mass spectrometry-based colony identification with a traditional CFU assay. This method was assessed using defined consortia made from up to eight bacterial strains. Among four replicate batches of an eight-strain mixture, observed values differed from expected values by less than 0.4 log10 CFU among all strains measured (range of differences, -0.318 to + 0.267). The average difference between observed and expected values was + 0.0308 log10 CFU, with 95% limits of agreement from -0.347 to 0.408 (Bland-Altman analysis). To estimate precision, a single batch of eight-strain mixture was assayed in triplicate by three different users, for a total of nine measurements. Pooled standard deviation values ranged from 0.067 to 0.195 log10 CFU for the eight strains measured, and user averages did not differ significantly. Leveraging emerging mass-spectrometry-based colony identification tools, a novel method for simultaneous enumeration and identification of viable bacteria from mixed-strain consortia was developed and tested. This study demonstrates the potential for this approach to generate accurate and consistent measurements of up to eight bacterial strains simultaneously and may provide a flexible platform for future refinements and modifications. KEY POINTS: • Enumeration of live biotherapeutics is essential for product quality and safety. • Conventional CFU counting may not differentiate between strains in microbial products. • This approach was developed for direct enumeration of mixed bacterial strains simultaneously.

Evaluating the Accuracy of 2020 Census Block-Level Estimates in California.

In this study, we provide an assessment of data accuracy from the 2020 Census. We compare block-level population totals from a sample of 173 census blocks in California across three sources: (1) the 2020 Census, which has been infused with error to protect respondent confidentiality; (2) the California Neighborhoods Count, the first independent enumeration survey of census blocks; and (3) projections based on the 2010 Census and subsequent American Community Surveys. We find that, on average, total population counts provided by the U.S. Census Bureau at the block level for the 2020 Census are not biased in any consistent direction. However, subpopulation totals defined by age, race, and ethnicity are highly variable. Additionally, we find that inconsistencies across the three sources are amplified in large blocks defined in terms of land area or by total housing units, blocks in suburban areas, and blocks that lack broadband access.

Pharmacoinformatics and Breed-Based De Novo Hybridization Studies to Develop New Neuraminidase Inhibitors as Potential Anti-Influenza Agents.

Influenza represents a profoundly transmissible viral ailment primarily afflicting the respiratory system. Neuraminidase inhibitors constitute a class of antiviral therapeutics employed in the management of influenza. These inhibitors impede the liberation of the viral neuraminidase protein, thereby impeding viral dissemination from the infected cell to host cells. As such, neuraminidase has emerged as a pivotal target for mitigating influenza and its associated complications. Here, we apply a de novo hybridization approach based on a breed-centric methodology to elucidate novel neuraminidase inhibitors. The breed technique amalgamates established ligand frameworks with the shared target, neuraminidase, resulting in innovative inhibitor constructs. Molecular docking analysis revealed that the seven synthesized breed molecules (designated Breeds 1-7) formed more robust complexes with the neuraminidase receptor than conventional clinical neuraminidase inhibitors such as zanamivir, oseltamivir, and peramivir. Pharmacokinetic evaluations of the seven breed molecules (Breeds 1-7) demonstrated favorable bioavailability and optimal permeability, all falling within the specified parameters for human application. Molecular dynamics simulations spanning 100 nanoseconds corroborated the stability of these breed molecules within the active site of neuraminidase, shedding light on their structural dynamics. Binding energy assessments, which were conducted through MM-PBSA analysis, substantiated the enduring complexes formed by the seven types of molecules and the neuraminidase receptor. Last, the investigation employed a reaction-based enumeration technique to ascertain the synthetic pathways for the synthesis of the seven breed molecules.

(p,q)-biclique counting and enumeration for large sparse bipartite graphs.

In this paper, we study the problem of (p, q)-biclique counting and enumeration for large sparse bipartite graphs. Given a bipartite graph G=(U,V,E) and two integer parameters p and q, we aim to efficiently count and enumerate all (p, q)-bicliques in G, where a (p, q)-biclique B(L, R) is a complete subgraph of G with L⊆U, R⊆V, |L|=p, and |R|=q. The problem of (p, q)-biclique counting and enumeration has many applications, such as graph neural network information aggregation, densest subgraph detection, and cohesive subgroup analysis. Despite the wide range of applications, to the best of our knowledge, we note that there is no efficient and scalable solution to this problem in the literature . This problem is computationally challenging, due to the worst-case exponential number of (p, q)-bicliques. In this paper, we propose a competitive branch-and-bound baseline method, namely BCList, which explores the search space in a depth-first manner, together with a variety of pruning techniques. Although BCList offers a useful computation framework to our problem, its worst-case time complexity is exponential to p+q. To alleviate this, we propose an advanced approach, called BCList++. Particularly, BCList++ applies a layer-based exploring strategy to enumerate (p, q)-bicliques by anchoring the search on either U or V only, which has a worst-case time complexity exponential to either p or q only. Consequently, a vital task is to choose a layer with the least computation cost. To this end, we develop a cost model, which is built upon an unbiased estimator for the density of 2-hop graph induced by U or V. To improve computation efficiency, BCList++ exploits pre-allocated arrays and vertex labeling techniques such that the frequent subgraph creating operations can be substituted by array element switching operations. We conduct extensive experiments on 16 real-life datasets, and the experimental results demonstrate that BCList++ significantly outperforms the baseline methods by up to 3 orders of magnitude. We show via a case study that (p, q)-bicliques optimizes the efficiency of graph neural networks. In this paper, we extend our techniques to count and enumerate (p, q)-bicliques on uncertain bipartite graphs. An efficient method IUBCList is developed on the top of BCList++, together with a couple of pruning techniques, including common neighbor refinement and search branch early termination, to discard unpromising uncertain (p, q)-bicliques early. The experimental results demonstrate that IUBCList significantly outperforms the baseline method by up to 2 orders of magnitude.

Analysis of factors related to cognitive impairment in a community-based, complete enumeration survey in Japan: the Nakayama study.

BACKGROUND: The number of patients with cognitive disorders is rapidly increasing in the world, becoming not only a medical problem, but also a social problem. There have been many reports that various factors are associated with cognitive dysfunction, but the factors have not yet been fully identified. This was a community-based complete enumeration study which aimed to identify risk and protective factors for dementia. METHODS: The first phase included all residents aged 65 years or older in a town in Japan. They completed many examinations, such as living conditions questionnaires, physical examination, Mini-Mental State Examination, and brain magnetic resonance imaging. The participants with suspected cognitive impairment underwent additional examinations for detailed evaluation in the second phase. Statistical analysis was performed to identify risk and protective factors for dementia after all participants were diagnosed. RESULTS: There were 927 participants in the baseline evaluation; 611 (65.9%) were healthy, 165 (17.8%) had mild cognitive impairment (MCI), and 151 (16.3%) had dementia. The age-standardised prevalence of dementia was 9.5%. Statistical analyses for amnestic MCI and Alzheimer's disease showed that risk factors for cognitive decline were diabetes mellitus, low activities of daily living, and living alone, and that protective factors were history of exercise and drinking habit. CONCLUSION: The present findings suggest that several lifestyle-related diseases and factors are associated with cognitive decline. These results support similar findings from previous studies and will be helpful for preventing dementia in the future.

Meeting the COVID challenge: Optimizing vCD34+ in cryopreserved HPC samples for implementation of an external QA Program.

BACKGROUND: The COVID-19 pandemic has forced a fundamental change in the global procurement of allogeneic hematopoietic progenitor cells (HPCs) for transplantation. To better meet the emergent challenges of transporting cryopreserved allogeneic HPC during pandemics, there is an urgent need for External Quality Assurance (EQA) programs to evaluate reproducibility and harmonization of viable CD34+ cell (vCD34+) HPC enumeration, as the current EQA programs are unsuitable for analysis of vCD34+. The cost-effective distribution of HPC cryopreserved reference samples (CRSs) with acceptable reproducibility and specificity is key to the success of a vCD34+ EQA program. METHODS: Cryopreserved HPC samples (n = 11) were either stored on dry ice for 1 to 4 days or for 1 day followed by liquid nitrogen (LN) storage for 1 to 3 days to assess optimal conditions for vCD34+ EQA. Flow cytometric enumeration of vCD34+ HPCs was performed using a single platform assay combined with 7-AAD viability dye exclusion. The optimum transportation condition was validated in pilot and multicenter national studies (n = 12). RESULTS: A combination of 1 day on dry ice followed by LN storage stabilized viability compared with continuous storage on dry ice. This study demonstrates that dispatch of CRSs on dry ice to recipient centers across a distance of ≤4000 km within 26 h, followed by LN storage, resulted in reproducible intercenter vCD34+ enumeration. The estimated cost of safer and more convenient dry ice delivery is >20-fold lower than that of LN. CONCLUSION: This approach can form the basis for economically and scientifically acceptable distribution of CRSs for external vCD34+ EQA.

Direct separation and enumeration of CTCs in viscous blood based on co-flow microchannel with tunable shear rate: a proof-of-principle study.

Circulating tumor cells (CTCs), which have extremely low density in whole blood, are an important indicator of primary tumor metastasis. Isolation and enumeration of these cells are critical for clinical applications. Separation of CTCs from massive blood cells without labeling and addition of synthetic polymers is challenging. Herein, a novel well-defined co-flow microfluidic device is presented and used to separate CTCs in viscous blood by applying both inertial and viscoelastic forces. Diluted blood without any synthetic polymer and buffer solution were used as viscoelastic fluid and Newtonian fluid, respectively, and they were co-flowed in the designed chip to form a sheath flow. The co-flow system provides the function of particle pre-focusing and creates a tunable shear rate region at the interface to adjust the migration of particles or cells from the sample solution to the buffer solution. Successful separation of CTCs from viscous blood was demonstrated and enumeration was also conducted by image recognition after separation. The statistical results indicated that a recovery rate of cancer cells greater than 87% was obtained using the developed method, which proved that the direct separation of CTCs from diluted blood can be achieved without the addition of any synthetic polymer to prepare viscoelastic fluid. This method holds great promise for the separation of cells in viscous biological fluid without either complicated channel structures or the addition of synthetic polymers.

Circulating tumor cells: A step toward precision medicine in hepatocellular carcinoma.

Serum alpha-fetoprotein and radiologic imaging are the most commonly used tests for early diagnosis and dynamic monitoring of treatment response in hepatocellular carcinoma (HCC). However, the accuracy of these tests is limited, and they may not reflect the underlying biology of the tumor. Thus, developing highly accurate novel HCC biomarkers reflecting tumor biology is a clinically unmet need. Circulating tumor cells (CTCs) have long been proposed as a noninvasive biomarker in clinical oncology. Most CTC assays utilize immunoaffinity-based, size-based, and/or enrichment-free mechanisms followed by immunocytochemical staining to characterize CTCs. The prognostic value of HCC CTC enumeration has been extensively validated. Subsets of CTCs expressing mesenchymal markers are also reported to have clinical significance. In addition, researchers have been devoting their efforts to molecular characterizations of CTCs (e.g. genetics and transcriptomics) as molecular profiling can offer a more accurate readout and provide biological insights. As new molecular profiling techniques, such as digital polymerase chain reaction, are developed to detect minimal amounts of DNA/RNA, several research groups have established HCC CTC digital scoring systems to quantify clinically relevant gene panels. Given the versatility of CTCs to provide intact molecular and functional data that reflects the underlying tumor, CTCs have great potential as a noninvasive biomarker in HCC. Large-scale, prospective studies for HCC CTCs with a standardized protocol are necessary for successful clinical translation.