Low permeability heterogeneous gas storage is one of the important types of gas storage currently, and clarifying its microscopic gas water movement laws is crucial for improving the operational efficiency of gas storage. This study investigated the impact mechanism of gas injection and production rates on the gas-water movement behaviors at the pore-scale in low-permeability heterogeneous gas storages. To achieve this, a series of microfluidics experiments was conducted using large-scale chips under different injection and production rate conditions. The dynamic evolution of gas-water distributions, migration pathways, and gas-bearing pore spaces was recorded and quantitatively analyzed. The results indicate that slow injection combined with fast production significantly improved the utilization efficiency of gas-bearing pore spaces. Slow injection promoted more uniform gas displacement and reduced fingering, while the fast production suppressed water occupation in pore throats and mitigated water-locking effects. Conversely, fast injection and slow production led to rapid early gas invasion but increased the formation of water-locked capillary valves, resulted in a higher proportion of bound pore-throats and reduced effective gas storage spaces during subsequent rounds. Microscopic observations showed that repeated injection–production processes progressively increase water-locked throats, limiting further gas migration. Parameter analysis based on capillary number demonstrated that the slow injection/fast production strategies favored the expansion of dynamic gas-bearing boundaries and lowered bound-throat ratios. Furthermore, pore-scale transport mechanisms are discussed using single pore-throat models, highlighted the roles of wettability, corner flow, and capillary resistance in controlling gas advance and retreat behaviors. The findings providing valuable insights into optimizing injection-production strategies for low-permeability heterogeneous gas storages, and offering theoretical support for improving gas storage efficiency and reducing water-locking effects.

The construction of underground gas storage (UGS) in a large-scale low-permeability lithologic gas reservoir presents an immense engineering challenge. Under the context of UGS, research on structural characteristics and storage capacity at the microscopic scale is insufficient, making it difficult to provide effective support for the engineering scheme. In this study, the microscopic storage spaces of a typical lithologic gas reservoir (i.e., YL block in the Ordos Basin) are comprehensively analyzed through experimental techniques (represented by computed tomography scanning), digital core analysis, and fractal analysis. Furthermore, the feasibility of UGS construction is examined. The results demonstrate that the large-scale low-permeability lithologic gas reservoir exhibits significant zonal heterogeneity in its microscopic structural characteristics at both morphological and statistical levels. Specifically, the microscopic storage spaces of the core zone within the YL block are notably higher than those in the transition and periphery zones, characterized by larger aperture, less tortuous, higher aggregation and connectivity. Consequently, the core zone provides adequate storage capacity and injection-extraction capability for large-scale underground storage of natural gas. In contrast, the transition and periphery zones exhibit inferior microstructural, storage, and flow properties, which are not suitable for rapid injection and production. However, these zones show a fairly strong lateral sealing capability, which can be utilized as a monitoring area to evaluate UGS integrity. These findings indicate that the reservoir's microstructural features meet the essential requirements of storage capacity, injection-extraction capability, and lateral sealing property for UGS construction. Based on this understanding, a series of zone-differentiated UGS engineering suggestions are proposed, including zonal function specification, well type selection, well deployment scheme, and management of old wells. These findings can provide valuable insights for the assessment and implementation of UGS projects from such gas reservoirs.

Hydraulic fracturing is essential for developing unconventional oil and gas resources, yet the imbibition process within fracture-matrix systems during shut-in remains unclear. In this study, X-ray computed tomography was employed to reconstruct the pore structure of tight sandstone with embedded tree-shaped fractures, generating fractured porous media. A novel self-adjusting pressure boundary condition was proposed by setting two virtual buffers at the inlet and outlet to simulate pressure diffusion. Pore-scale simulations were conducted to examine the effects of initial pressure difference, wettability, viscosity ratio, and interfacial tension on dynamic imbibition. Results show that pressure release from the inlet to the outlet buffer gradually reduces the driving force for flow, shifting pore-scale displacement from viscous- to capillary-dominated flow. Two distinct local imbibition behaviors, including counter-current and forced imbibition, were simultaneously observed in the fractured medium, and their kinetic mechanisms were analyzed. Increasing the initial pressure difference from 2 & times; 105 Pa to 5 & times; 105 Pa enhanced water infiltration into fractures and matrix, raising water saturation from 0.30 to 0.62. Although fluid properties have limited influence on the overall water saturation profile, they significantly affect local imbibition processes, thereby altering oil-water distribution patterns, which may have notable implications for subsequent depressurization-induced production.

Ethanol cosolvent flushing is an effective in-situ technique for enhancing remediation of groundwater contaminated by dense non-aqueous phase liquids (DNAPLs). The remediation efficiency critically depends on the dissolution and mass-transfer performance of residual DNAPL. However, due to the intrinsic opacity of aquifer materials, the sand-column systems commonly used in flushing experiments provide limited insight into the dynamic behaviors of DNAPL and the underlying mechanisms governing its removal. In this study, high-resolution X-ray micro-computed tomography was employed to visualize and quantify the dissolution and mass transfer of residual DNAPL in porous media under various flow rates and ethanol contents. Results indicate that increasing flow rate enhances both dissolution rates and mass transfer coefficients. The effect of ethanol content is non-linear: dissolution rates increase gradually at low contents but rise sharply when ethanol exceeds ∼30%, consistent with solubility trends. Ethanol also alters residual DNAPL morphology and flow paths, enlarging interfacial area and counteracting the expected increase in mass transfer resistance. Moreover, local heterogeneity in mass transfer triggers dynamic interfacial processes such as receding, snap-off, and breakup, which influence dissolution behavior. Furthermore, a new empirical correlation was then developed, integrating saturation, flow rate, and solubility, which can accurately predict mass transfer coefficients during both water and cosolvent flushing. This work elucidates key pore-scale mechanisms controlling residual NAPL mass transfer and proposes a universal model for solubilization-enhanced remediation.

Interphase mass transfer between residual phase and flowing water in porous media is a critical process in subsurface engineering applications, yet its dependence on complex subsurface wettability remains poorly understood. In this study, fractional-wettability porous media were constructed by mixing different proportions of water-wet and neutral-wet grains. High-resolution micro-X-ray computed tomography enabled direct pore-scale visualization of nonaqueous phase liquid (NAPL) morphology and dissolution dynamics. The result indicates that wettability heterogeneity controls interphase mass transfer primarily through its impact on the morphology of residual phase. Fully water-wet media favor small singlets and ganglia that facilitate uniform size distribution and effective contact with flowing water, whereas neutral-wet grains induce large, spanning clusters that lead to flow bypassing. Pore-scale interface dynamics of individual NAPL blobs indicates that interphase mass transfer is nonuniform. The NAPL-water interface comprises the terminal interface with positive curvature, which contacts the main flow and enhances interphase mass transfer, and a negative surface constrained by solid walls, where thin-film contact limits the transfer rate. Wettability heterogeneity substantially alters the fraction of terminal interfaces within the total NAPL-water interfacial area, yielding average values of 0.48 in water-wet media and 0.29 in neutral-wet media. After decoupling interfacial area effects, the intrinsic liquid-liquid mass transfer coefficient was found to be on the order of 10-7 m/s at Re approximate to 0.09. During dissolution, as residual saturation decreases, the coefficient initially rises, enters a plateau region, and eventually declines at low saturations, reflecting the evolving morphology of residual phase and its control on interphase mass transfer.

Immiscible displacement in porous media is governed by pore-scale interfacial dynamics and strongly modulated by wettability heterogeneity. Pore-scale observations indicate that wettability can depend on pore size, yet the consequences for invasion patterns remain unclear. Here we perform direct numerical simulations of immiscible displacement under four wettability conditions, focusing on two size-dependent mixed-wet distributions: (i) a binary distribution with an abrupt weakly hydrophilic–strongly hydrophobic transition and (ii) a continuum distribution with a smooth wettability gradient. In both mixed-wet cases, larger pores are assigned more hydrophobic contact angles; uniformly weakly hydrophilic and uniformly strongly hydrophobic conditions serve as references. At Ca=2.0×10−5 and viscosity ratio M=0.02, the continuum distribution develops the narrowest preferential pathways and the highest residual saturation, whereas the binary distribution yields more compact invasion and lower residual saturation. Mechanistic analysis of representative pore-filling events shows that the binary distribution promotes contact-line pinning at sharp wettability transitions, suppressing post-Haines-jump relaxation and limiting fingering, while the continuum distribution retains partial mobility that sustains preferential invasion. Breakthrough residual saturation decreases with decreasing Ca in the binary case but is nearly independent of Ca in the continuum case, and decreases with increasing M in all cases. Reversing the pore-size wettability trend suppresses fingering most clearly at intermediate times and yields lower residual saturation in the binary case than in the corresponding continuum case. These results show that the functional form of pore-size-dependent wettability strongly controls invasion pathways and residual trapping. Predictive models of subsurface immiscible displacement should therefore represent pore-scale wettability correlations, rather than only their average strength or overall trend.

Wettability governs two-phase flow in porous media, thereby directly affecting the security of CO2 geological sequestration and oil recovery efficiency, yet the role of mixed-wettability remains poorly understood. This study investigates the dynamics of capillary trapping using a real-rock microfluidic chip that preserves the authentic pore structures and mineral composition of natural rock. By employing high-resolution in situ imaging, we compared waterflooding behaviors under water-wet and stearic acid-induced mixed-wet conditions across a range of capillary numbers. In unaged water-wet media, residual oil is primarily trapped as isolated singlets due to the snap-off events driven by the wetting films and corner flows. High injection rates were found to suppress this mechanism by restricting the film development time, thereby significantly reducing residual oil saturation. Conversely, wettability alteration fundamentally shifts the trapping regime. In mixed-wet systems, residual oil saturation increased significantly, forming extensive multi-pore clusters anchored to oil-wet regions. This enhanced trapping results from the inability of water to form stable films on oil-wet walls, which suppresses snap-off events and promotes macroscopic bypassing. Furthermore, a direct correlation between local wettability distribution and trapping mechanics was established, identifying four distinct behaviors in mixed-wet porous media: classical snap-off, local bypassing in asymmetric pores, ganglion trapping, and macroscopic bypassing across continuous oil-wet zones. These findings demonstrate that the spatial distribution of wettability governs the transition from film-driven snap-off to wettability-induced bypassing, serving as the dominant factor controlling residual oil morphology in complex porous media.

This study investigates pore-scale flow dynamics and pressure-energy accumulation mechanisms during prefracturing injection in ultra-low permeability porous media, explicitly treating the injection process as a transient pore-scale energy storage process rather than a purely hydraulic operation. Pore-scale two-phase flow simulations based on a volume-of-fluid (VOF) framework are conducted under closed-boundary conditions to quantify pressure redistribution and energy accumulation behavior prior to fracture initiation. Results: reveal that capillary-driven flow instabilities, including upstream backflow, channel reconnection, and oscillatory interface motion at pore-throat discontinuities, govern localized pressure buildup and temporary energy storage. Energy analysis shows that the total stored energy increases monotonically and is consistently dominated by the oil-phase contribution. Despite variations in injection rate, similar energy accumulation trends are observed, indicating that injected volume rather than injection rate primarily controls pore-scale energy storage under capillary-dominated conditions. In contrast, wettability exerts a strong influence on invasion pathways and energy distribution, with water-wet conditions facilitating deeper penetration and more effective pressure transmission, while oil-wet conditions promote stagnation, backflow, and residual trapping. Variations in oil-water viscosity ratio have only a secondary effect on interface morphology and energy evolution. These findings demonstrate that pore-scale invasion dynamics and wettability, rather than injection rate alone, dominate pressure response and energy storage during pre-fracturing injection, providing new mechanistic insights for optimizing injection strategies in unconventional reservoirs.

The unique operational environment of AI-driven data centers exacerbates high energy consumption, a large portion of which is attributed to electronic cooling systems. Developing energy-efficient thermal management solutions is critical for establishing green data centers. Thermosyphons have emerged as highly effective passive cooling devices, offering a promising approach. However, the electronic miniaturization leads to the development of compact thermosyphons, influencing their thermal performance. Regarding structural design constraints, surface modification presents a viable optimization strategy. This study experimentally investigated the boiling heat transfer enhancement in a 10 mm-height compact thermosyphon featuring modified surfaces. Wettability hybrid and gradient surfaces (hydrophobic 120 degrees- hydrophilic 30 degrees/75 degrees) were fabricated using an ultraviolet laser etching technique. The effects of surface modification on bubble dynamics were explored. The thermal performance enhancement was evaluated by the heat transfer coefficient, overall thermal resistance and dimensionless parameters. The experimental results illustrate that, compared to smooth surface case, the 120 degrees- 75 degrees modified surfaces exhibit limited thermal performance improvement due to bubble expansion effects. Both 120 degrees-30 degrees modified surfaces featured larger contact angle difference demonstrate superior heat transfer in the compact thermosyphon. Specifically, the 120 degrees-30 degrees wettability hybrid surface promotes bubble nucleation and prevents bubble expansion on the condensation surface, while the 120 degrees-30 degrees wettability gradient surface enhances liquid replenishment and bubble detachment. Thermal performance analysis shows that the 120 degrees-30 degrees modified surfaces achieve maximum heat transfer enhancements of 2.96 and 3.29 times that of the smooth surface. This research aims to provide design guidance for optimization of integrated heat sinks in electronic cooling fields.

The dynamic evolution of gas–solid–liquid interfaces exert a critical influence on reactive transport in subsurface engineering, including geological CO2 sequestration and unconventional energy development. However, the microscale interfacial processes by which reaction-induced bubbles regulate mineral dissolution remain poorly understood, owing to the challenge on visualizing in the opaque natural rock. In this work, emerging mineral microfluidic platform was employed to directly visualize and quantify the coupling between CO2 bubble generation and single calcite dissolution. Three distinct dissolution regimes were identified depending on flow rate and acid concentration, namely bubble-free, bubble-shielding, and bubble-sealing dissolution. At low acid concentration, dissolution proceeds smoothly without bubble formation, and the effective reaction rate increases with flow rate following a power-law trend. Once a critical acid concentration is exceeded, CO2 bubbles nucleate and grow on the reactive surface, sharply reducing the effective reactive area. Acid transport becomes confined to narrow channels surrounding the bubbles, suggesting a progressive shift from convection-dominated to diffusion-dominated mass transfer. Bubble shielding significantly decreases the effective reaction rate by 81.5

Background:To control HIV-1 viral load and maintain optimal immune function, albuvirtide (ABT) combined with antiretroviral therapy (ART) is used in designated AIDS hospitals in China. However, the standard weekly 320-mg ABT injection regimen poses adherence-related challenges. This study explored the safety and virological outcomes of a modified ABT intensification regimen in people with HIV. Methods:This single-arm, uncontrolled prospective cohort enrolled 20 people with HIV who received intravenous ABT 320 mg for three consecutive days every four weeks while continuing their baseline ART regimens. Peripheral blood samples were collected to measure alanine aminotransferase (ALT), aspartate aminotransferase (AST), triglyceride (TAG), cholesterol, CD4+ T-cell counts, HIV-RNA levels, and plasma ABT concentrations. Results:No adverse events considered related to ABT were observed in this cohort. Pharmacokinetic analysis demonstrated sustained therapeutic exposure, with mean trough concentrations of 4.71 mg/L, 3.83 mg/L, and 4.53 mg/L at Weeks 4, 8, and 12, respectively, each exceeding 50-fold the IC9 0. During follow-up, mean HIV-RNA decreased from 1.65 log10 copies/mL at baseline to 0.59 log10 copies/mL at Week 12 (mean reduction: 1.06 log1 0 copies/mL; P = 0.0007). Among patients with baseline HIV-RNA above 20 copies/mL, the viral load decreased from 2.55 to 1.43 log10 copies/mL. Conclusion:Preliminary results demonstrate that 4-week ABT antiviral intensification regimen is feasible and exhibits potential antiviral activity.

A comprehensive understanding of spontaneous imbibition mechanisms in fractured porous media is crucial for optimizing the development and sustainable utilization of geo-energy resources. This study employs advanced pore-scale simulation techniques to investigate counter-current spontaneous imbibition behavior in natural fractured porous media. Through digital rock physics technology, we accurately reconstructed both matrix and fracture structures from low-permeability sandstone cores. The study revealed two distinct snap-off phenomena in natural fractured porous media: (1) fracture-dominated snap-off, which plays a pivotal role in oil–water phase redistribution, and (2) matrix-dominated snap-off, which exhibits relatively minor effects on fluid phase redistribution. Three key geometric parameters significantly influence the spontaneous imbibition process in natural fractured porous media: the pore–throat ratio of matrix, the throat size distribution at the matrix–fracture junctions, and the fracture architecture. The relatively large pore–throat ratio in natural fractured porous media significantly inhibits the spontaneous imbibition capacity of water, consequently leading to reduced oil recovery efficiency. In addition, the systematic analysis focused on three critical controlling factors of fluid properties: wettability characteristics, viscosity ratio, and interfacial tension, with particular emphasis on their influence on interaction mechanism between fracture and matrix. Increasing interfacial tension or decreasing contact angle can promote deeper water penetration into the matrix, thereby improving oil recovery efficiency. Furthermore, our findings indicate that higher viscosity ratios enhance non-wetting phase mobility, facilitating more efficient spontaneous imbibition processes. These findings provide valuable insights for optimizing water injection strategies in unconventional reservoirs, such as oil shale formations.

We report a case of an adult male with a history of recurrent cough and chest pain for five years. Mycobacterium arosiense was identified in his alveolar lavage fluid, and whole-exome sequencing revealed heterozygosity for CD209 in this patient. After 17 months of combined antibiotic therapy, the patient recovered completely.

To gain an in-depth understanding of the microscopic flow characteristics and residual oil migration during viscoelastic particle flooding， this study established a pore-scale numerical model for viscoelastic particle-water-oil three-phase flow based on digital core. By tracking the dynamic evolution of the oil and water distribution field and analyzing the spatial structure changes of flow paths， the characteristics and evolution laws of flow paths during viscoelastic particle flooding were systematically revealed. Simulation results indicate that compared with conventional water flooding， dynamically branching flow paths are formed in the porous medium due to the plugging effect on the pore channels or increasing the flow resistance of viscoelastic particles in the pore channels. The evolution of these flow paths exhibits three distinct stages： path reconstruction， where the number of flow paths rapidly decreases from 60 to 25； path optimization， where the standard deviation of the weight increases from 5.00 to 10.00， accompanied by a 22% reduction in residual oil volume fraction； and dynamic equilibrium， where the number of flow paths remains around 15， and the residual oil volume fraction further decreases from 0.26 to 0.13. Quantitative analysis further shows that the resistance regulation and pressure fluctuations during viscoelastic particle flooding prompt fluids to enter previously unmobilized regions， improving displacement efficiency. The volume fraction of mobilized zones increases to 0.50-0.70， and the overall oil recovery is increased by 35% compared to the end of water flooding. The research reveals the dynamic evolution characteristics of the flow path during the viscoelastic particle flooding and clarifies the relationship between its microscopic regulatory mechanism and the macroscopic oil recovery efficiency.

ABSTRACTBased on large eddy simulations, intermittent airflow within an urban street canyon was simulated. The practice of time‐varying inflow conditions (TVIC) required a time series of inflow wind velocity, which could be collected on a varying curve of the moving averaged measured data. The influences of the time interval of the wind series and the varying trend (or molded line) between adjacent data on airflow within the street canyon were analyzed. The results showed that TVIC would result in larger average wind velocity and turbulence intensity than that simulated under steady inflow conditions (SIC). The simulated total vertical air exchanges under TVIC would be one order of magnitude higher than that simulated under SIC. Airflow characteristics within street canyons were influenced by the varying trends and the time intervals of the time‐series inflow wind. Average vertical wind velocity and turbulent kinetic energy (TKE) simulated under the stepped varying trend was higher than that under the jagged varying trend. The shorter the time interval, the larger the TKE within the street canyon. Vertical air exchanges induced by turbulence (ACH′) at the roof level simulated under the stepped molded lines were twice that of the jagged molded line. Under the time interval of 30 s, the ACH′ was significantly increased, which was 2.558 times that simulated with a time interval of 1 min. Thus, the suggested practical approach for time‐varying inflow simulations is to obtain time‐series wind data with a time interval of 1 min or less, and the linearly molded line would be critical; for larger time intervals, reasonable molded lines would be required.

Understanding CO2 dissolution in porous media is crucial for assessing the efficiency and safety of geological carbon storage. While empirical power-law correlations can capture the effect of flow velocity on mass transfer, they often overlook the influence of bubble morphology. In this study, we employ pore-scale direct numerical simulations to investigate the dissolution process of trapped CO2 bubbles in porous media with two typical configurations: singlets and clusters. Key transport parameters, including interfacial area, flow field distribution, local P & eacute;clet number, and concentration distribution, were analyzed under different flow rates and saturations to evaluate performance. Results reveal a non-monotonic trend in mass transfer coefficients for singlet bubbles with increasing saturation-initially rising and then declining. At low saturations (<10%), increased gas-liquid interfacial area and enhanced pore-scale water velocity strengthen convective transport and promote dissolution. In contrast, at high saturations (>15%), the development of extensive stagnant flow regions suppresses local concentration gradients, resulting in diffusion-dominated transport and reduced overall mass transfer rate. A comparison between bubble morphologies shows that singlet bubbles dissolve significantly faster than clusters, not only due to greater interfacial area but also higher overall mass transfer coefficients. Moreover, for singlet bubbles, a consistent power-law relationship between Sherwood and Reynolds numbers is observed across all conditions, with the scaling exponent strongly depending the gas saturation and bubble configuration. These findings emphasize the critical role of bubble morphology in CO2 dissolution and provide quantitative data for improving continuum-scale reservoir models.

Mineral dissolution plays a fundamental role in key subsurface processes, involving complex coupling among fluid flow, mass transport, and geochemical reactions. In this study, we present microfluidic chips fabricated from natural carbonate rocks using laser micromachining with a layer-by-layer strategy. The chips are optically transparent with a non-reactive basal surface, preserving mineralogical heterogeneity while enabling real-time, high-resolution visualization of dissolution. Four representative carbonate rocks-calcite, mixed-mineral, dolomite, and multi-mineral-were investigated to assess the influence of mineral composition and heterogeneity on dissolution dynamics. Results show that mineral composition strongly governs dissolution patterns: calcite rocks develop wormhole structures, while low-reactive dolomite rocks exhibit compact patterns. Notably, in mixed carbonate rocks containing both calcite and dolomite, even when calcite accounts for up to 33% of the composition, the overall dissolution dynamics-including the effective reaction rate-are mainly dominated by dolomite. Micro-mechanistic analysis indicates that this occurs because the highly reactive calcite dissolves preferentially, creating new pores between the less reactive dolomite crystals. This limits the effective transport of reactive fluid, thereby suppressing further reaction. The remaining dolomite framework eventually dissolves completely, resulting in a compact dissolution pattern. In contrast, in multi-mineral rocks rich in quartz and clay, the inert minerals accumulate at the reaction front, forming a stable dissolution-altered layer, which macroscopically manifests as a negligible change in particle morphology. This study establishes a versatile experimental platform for investigating fluid-rock interactions in subsurface engineering and provides new insights into how mineral composition and heterogeneity influence dissolution behavior in natural rocks.

Background:Non-small cell lung cancer (NSCLC) patients with double-driver gene mutations are reported with poorer survival outcomes and reduced therapy responses compared to single-mutant (SM) patients. While substantial progress has been made in treating cancers with single-driver gene alterations, the therapeutic implications and tumor microenvironment of double-mutant (DM) NSCLC remain largely unexplored because of its rarity and poor prognosis. This study aims to delineate the landscape of the immune microenvironment within DM NSCLC. Methods:We employed single-cell RNA sequencing (scRNA-seq) to generate a comprehensive transcriptomic atlas from 25 EGFR mutant NSCLC samples. To assess immune microenvironment changes over time, time-series scRNA-seq and multiplex immunofluorescence analyses were performed in two DM patients. In vitro, EGFR-mutant cell lines were constructed to assess potential therapeutic responses for future clinical applications. Results:The scRNA-seq platform scFocuSCOPE accurately identified and characterized rare, mutation-bearing cancer cells at the single-cell level. DM cancer cells exhibited a strong tendency toward angiogenesis, suggesting an invasive phenotype. DM patients had a more suppressed immune microenvironment, with fewer dysfunctional T lymphocytes. The observation of fewer immune cells and high programmed death ligand-1 (PD-L1) expression in DM cases, probably related to immune evasion and poorer prognosis. In one DM patient, PD-L1 expression remained unchanged after targeted therapy but decreased after immunotherapy. In vitro, EGFR/ERBB2 DM cells showed greater sensitivity to dual-targeted therapies than to single-agent treatments. Conclusions:scFocuSCOPE precisely delineated tumor heterogeneity and immune suppression in EGFR-DM NSCLC. The complex immune landscape of EGFR-DM tumors offers valuable insights for future mechanistic studies and personalized therapies.