ABSTRACT Multivalent ligand–receptor interactions underlie most forms of cell-cell communication, yet a general quantitative framework for “avidity” has remained elusive for over a century. Here, we derive closed-form expressions for signaling potency (EC 50 ) in multivalent systems directly from first principles, extending exact analytical models of ternary complex equilibria to account for receptor confinement at cell surfaces. These equations unify antibody-antigen and cytokine-receptor interactions under a common mathematical framework in which potency emerges as a function of binding constants and receptor density. In contrast to monovalent models, EC 50 is no longer equal to the dissociation constant (K d ), but instead reflects receptor-dependent avidity effects that vary across cellular contexts. We validate these predictions across biophysical measurements, in vitro binding and signaling assays, in vivo murine cytokine perturbation data, and human spatial transcriptomic datasets. The framework explains longstanding empirical observations, including enhanced antibody potency through avidity and asymmetric control of cytokine signaling by receptor subunits. By embedding these equations within a regression-compatible formulation, we enable inference of signaling drivers from single-cell and spatial transcriptomic data. This work establishes a mechanistic bridge between molecular binding, receptor context, and tissue-level signaling, providing a quantitative foundation for interpreting and modeling intercellular communication in health and disease.

As drug development costs continue to rise, there is a need to reframe how drug efficacy is evaluated in preclinical models to reduce the rate of false positives. The “valley of death” refers to the gap between bench research and clinical translation. In particular, oncology chemotherapies have the highest rate of drug failure compared to other drug classes. While there has been progress in overall cancer survival, some cancers and patient populations still have a poor prognosis. To bridge the gaps of drug failure and aid underserved patient populations, drug translation cannot be viewed as a purely linear process, but one in which continual refinement is used to create more efficacious drug candidates. Additionally, pharmacokinetic-pharmacodynamic modeling could prove instrumental in better understanding drug efficacy in cellular models and in evaluating clinical translation potential with greater accuracy. The path forward in clinical pharmacology is to view drug development and efficacy as a dynamic process rather than a purely linear fashion. This review discusses traditional pharmacodynamic and pharmacokinetic evaluation methods, as well as pharmacokinetic-pharmacodynamic models of tumor growth inhibition.

Abstract Triple-Negative Breast Cancer (TNBC) presents a significant clinical challenge due to its heterogeneity and lack of targeted treatment options, with chemotherapy and immunotherapy combinations currently serving as the main therapeutic strategy. Efforts to address TNBC heterogeneity have largely focused on classifying intrinsic cancer subtypes based on differential tumor mRNA expression, a strategy that has proven effective in hormone receptor-positive breast cancers but has yet to yield a clinically useful predictor of survival or treatment response in TNBC. We hypothesize that both the intrinsic characteristics of TNBC and the surrounding immune microenvironment influence treatment outcomes and that immune cell infiltration affects TNBC subtype classification and response variability. To explore this hypothesis, we compared the predictive and prognostic capabilities of cancer subtype-based (TNBC-type) gene signatures and immune cell deconvolution methods (CIBERSORT) within the same TNBC datasets. We found that immune cell abundance outperformed TNBC subtype-signatures and multicellular immune cell aggregates showed the highest performance of all. More specifically, aggregate immune cells associated with tertiary lymphoid structures and tumor associated macrophages/monocytes demonstrated statistically significant predictive value. These findings were confirmed in an independent cohort of 67 TNBC patients treated with neoadjuvant chemotherapy. Further, single-cell RNA sequencing analysis revealed that the predictive power of cancer-subtype could be partially explained by immune- and stromal features. Examination of single-cell resolution spatial transcriptomic data confirmed presence of TLS-like, TAM- and cancer-stromal niches within TNBC biopsy samples that were associated with treatment response. Overall, our results highlight that immune cell aggregates, which capture the spatial organization of the TME, outperform cell-type specific gene signatures in predicting TNBC outcomes. Our novel approach provides a robust framework for interpreting spatial relationships in bulk RNA-seq data, offering a pathway for reconciling past data with current advancements in spatial profiling technologies. This work paves the way for future studies to leverage the multi-cellular complexity of TNBC, enhancing diagnostic precision and facilitating the development of therapies that strategically modulate the tumor microenvironment for improved anti-cancer responses.

Detailed Materials and Methods

Objectives: P-glycoprotein (Pgp) contributes to drug distribution (oral bioavailability[1], blood-brain barrier[2], stem cells and cancer chemoresistance[3]). Predicting drug access to tissues and cells is important to understand the drug response in-vivo. Unfortunately, Pgp specificity for FDA drugs is unclear due to technical variability in assays that quantify barrier ratios and enzyme kinetics (kcat/Km) using non-cellular experimental models[4]. Our research improves Pgp specificity scores by leveraging new Pgp expression and function datasets including drug screening data, Pgp expression data (cell lines, tissues) and barrier ratio datasets.Methods: Emerging big datasets (gene expression, drug response etc.) provide an opportunity to simultaneously measure Pgp quantity and function across tissues and cell lines. At the same time, physics-informed machine learning infers biophysical kinetic parameters from these datasets. Here, we experimentally and computationally integrate functional dataset information to better understand Pgp specificity. Pgp function is measured through barrier ratios (tissue and cells—Calcein AM[5]) and drug response data in cells. Pgp expression is quantified through RNAseq[6], proteomics and flow cytometry. We take these pre-existing datasets and create an experimental procedure that leverages both data types. We utilize multilinear regression as our machine learning approach. Our rationale is based on underlying Michaelis-Menten physics and enables us to make functional assays for patient-specific diagnostics.Results: We obtained consensus scores for Pgp specificity across 1,500 FDA approved drugs (PRISM[7]) and validated them experimentally in a subset of 76 substrates selected to represent top and bottom drugs for Pgp specificity. These scores can be used to calibrate clinical diagnostics (Pgp expression), and our experimental platform can be used to quantify Pgp function in clinical samples. Conclusions: Overall, we have developed a parallel computational and experimental procedure to estimate Pgp selectivity in live cells. The long term implications for this research include multidrug resistance diagnostics, tissue distribution predictions, drug-drug interaction predictions and transporter kinetics measurement (cellular vs protein) improvements.Citations: [1] Eur J Pharmacol., 2010, 627:92-98. [2] Cell, 1994, 77:491-502.[3] Endocr Relat Cancer., 2003, 10:43-73.[4] J Biol Chem., 2017, 292:15838-15848. [5] Eur J Pharm Sci., 2001, 12:205-214. [6] Nature, 2019, 569:503-508. [7] Nature Cancer, 2020, 1:235-248.

Supplementary Table 7. Transcriptomic analyses of the PDX models A. PDX interactome B. OncoTreat for drugs to PDX using DU145 cell perturbation data

Supplementary Table 3: Transcriptomic analyses of the human patient samples A. TCGA Interactome B. SU2C Interactome C. Protein activity TCGA D. Protein Activity SU2C

Supplementary Figures S1-S6 and figure legends.

Supplementary Table 6. OncoLoop summary table

Supplementary Table 2: Transcriptomic analyses of the GEMMs A. GEMMs Interactome B. Protein activity C. Cluster analyses D. Pathway analyses

Ternary-complex directed enzyme catalysis underlies a vast array of biological processes and several clinical therapies including growth hormones, interferon, and heparin. Recently, interest in ternary catalysis drugs has increased significantly with the rapid expansion of research new technologies such as bispecific antibodies and proteolysis targeting chimeras (PROTAC’s). Here, we derive a general model for ternary complex catalysis that defines the timescales of these diverse processes in familiar terms from classical enzyme theory. This was accomplished by solving for the maximum velocity (Vmax) and adapting an under-appreciated strategy within Michaels and Menten’s original publication: integration of the velocity equation. Critically, these equations are simple, conceptually accessible, and enables rapid estimation timescales that are consistent with a wide range of published literature. Finally, we have combined these equations with “big data” from new thermodynamic and kinetic databases to build interactive online tools that enable non-computational investigators to graphically simulate their own systems: • https://douglasslab.com/Btmax_kinetics/ Overall, this work is part of a general trend to reconceptualize pharmacodynamics from classical binding equilibria (e.g. Langmuir-Hill equation) to a kinetic processes with a characteristic timescale.

Supplementary Table 1: Phenotypic analysis of the GEMMs

Single-agent clinical response data for widely used chemotherapies have remained difficult to analyze because they are scattered across decades of print literature. We consolidated these sources into NCI1970-Meta, a 49,002-patient dataset covering 30 drugs across 18 cancers, enabling the first quantitative comparison of clinical outcomes, laboratory potency metrics, clinical exposure, and literature-derived biomarkers. Clinical patterns were strongly lineage-driven: cancer type explained far more ORR variability than drug identity (37.3% vs 15.6%; F = 13.43 vs 3.29). FDA approvals reflected these same patterns where ORR strongly predicted indication status (F = 98.3-104.1). In contrast, laboratory efficacy metrics did not track clinical activity. Raw in vitro AUC showed no association with ORR (R2 = 0.00; 95% CI: 0.00-0.01) and was dominated by drug identity rather than cancer lineage (85.2% vs 8.6%; F = 212.2 vs 21.6). Instead, AUC correlated with clinical exposure: unbound Cmax (R2 = 0.21 [0.21-0.34]) and therapeutic minimum concentrations (R2 = 0.69 [0.64-0.73]). This indicates that standard assay ranges capture exposure requirements rather than true efficacy. Normalizing potency by exposure restored the expected clinical relationships and resolved drug-specific anomalies such as gemcitabine. Biomarkers showed consistent behavior across clinical and laboratory settings. Among 314 biomarker-drug pairs, correlation directions were significantly conserved (R2 = 0.17 [0.10-0.25]; p = 2.8×10 −14 ). Literature-defined sensitivity and resistance annotations were enriched in vitro (OR = 3.7; p = 2.36×10 −8 ) and in the clinic (OR = 3.9; p = 6.52×10 −9 ), with stronger performance for correlations >0.1 (OR = 13.3 in vitro; OR = 8.4 clinically). Simple biomarker-sum models performed well across drugs, consistent with multi-pathway pseudo-first-order behavior. Overall, NCI1970-Meta provides a quantitative framework linking laboratory pharmacology to real-world clinical efficacy. Biological signal is reliably preserved within drugs, while cross-drug comparisons require explicit exposure normalization. This resource offers a statistical foundation for improving drug prioritization, biomarker development, and translational pharmacodynamic modeling.

Index of Supplementary Data

Figure S1: Genomic alterations in prostate cancer represented in the GEMMs (related to Fig. 2). Figure S2: Additional phenotypic analyses of the GEMMs (related to Fig. 2). Figure S3: Phenotypic analysis of allograft and organoid models (related to Fig. 2). Figure S4: Additional transcriptomic analyses of the GEMMs (related to Fig. 3). Figure S5: Analyses of AR activity in GEMMS (related to Fig. 3). Figure S6: Regulatory sub-networks of the GEMM clusters (related to Fig. 3). Figure S7. MR-match of prostate cancer cells lines to human PCa (related to Figs. 5). Figure S8. Drug perturbation protein activity profiles from DU145 cells (related to Figs. 5, 6, 7). Figure S9. LNCaP Pharmacotyping to patients and GEMMs (related to Figs. 5). Figure S10: Additional validation of drug candidates (related to Figs 6, 7). Figure S11: Additional validation of drug candidates (related to Figs 6).

Traditional laboratory methods quantify drug-efficacy based on dose-response metrics (e.g. IC50 or AUC), whereas clinical oncology defines drug-efficacy based on time-response metrics (e.g. progressive or stable disease). These measures often diverge, limiting translation between bench and bedside. We developed NCI1970-Meta, a harmonized dataset of 49,002 evaluable patients covering 30 drugs and 18 cancers, to provide the first large-scale benchmark linking in vitro potency, clinical pharmacokinetics, and Phase 2–3 outcomes data. Raw in vitro metrics values showed poor correlation with clinical response, systematically underestimating antimetabolites and overvaluing outliers such as gemcitabine. By contrast, normalizing potency to clinically achievable exposure (e.g. the maximum plasma concentration (Cmax)) restored concordance with trial outcomes. Biomarker analyses, which compare patients treated with the same drug, were largely robust to this dichotomy: 234 curated biomarkers showed concordance across clinical and in vitro datasets. In addition, curated transporter knockout data validated a parallel flux model of additive resistance, explaining the limited success of multi-drug resistance inhibitors for drug-transporter enzymes. These findings establish exposure-normalized potency as a simple, actionable framework for translational oncology. We recommend screening to the unbound Cmax and normalizing potency metrics by exposure to improve the fidelity of functional assays, reduce bias in AI models, and refine biomarker discovery. Beyond cytotoxic agents, NCI1970-Meta provides a general framework for aligning laboratory pharmacology with clinical outcomes. ### Competing Interest Statement The authors have declared no competing interest.

Supplementary Table 4: OncoMatch A. OncoMatch TCGA B. OncoMatch SU2C

Supplementary Table 5. Drug perturbation and OncoTreat analysis A. Summary of drugs and concentrations used B. Drug perturbation data for DU145 C. OncoTreat for drugs to SU2C patients using DU145 D. OncoTreat for drugs to GEMMs using DU145