Network Analyses of Brain Tumor Patients' Multiomic Data Reveals Pharmacological Opportunities to Alter Cell State Transitions.

Glioblastoma Multiforme (GBM) remains a particularly difficult cancer to treat, and survival outcomes remain poor. In addition to the lack of dedicated drug discovery programs for GBM, extensive intratumor heterogeneity and epigenetic plasticity related to cell-state transitions are major roadblocks to successful drug therapy in GBM. To study these phenomenon, publicly available snRNAseq and bulk RNAseq data from patient samples were used to categorize cells from patients into four cell states (i.e. phenotypes), namely: (i) neural progenitor-like (NPC-like), (ii) oligodendrocyte progenitor-like (OPC-like), (iii) astrocyte- like (AC-like), and (iv) mesenchymal-like (MES-like). Patients were subsequently grouped into subpopulations based on which cell-state was the most dominant in their respective tumor. By incorporating phosphoproteomic measurements from the same patients, a protein-protein interaction network (PPIN) was constructed for each cell state. These four-cell state PPINs were pooled to form a single Boolean network that was used for in silico protein knockout simulations to investigate mechanisms that either promote or prevent cell state transitions. Simulation results were input into a boosted tree machine learning model which predicted the cell states or phenotypes of GBM patients from an independent public data source, the Glioma Longitudinal Analysis (GLASS) Consortium. Combining the simulation results and the machine learning predictions, we generated hypotheses for clinically relevant causal mechanisms of cell state transitions. For example, the transcription factor TFAP2A can be seen to promote a transition from the NPC-like to the MES-like state. Such protein nodes and the associated signaling pathways provide potential drug targets that can be further tested in vitro and support cell state-directed (CSD) therapy.

Low-frequency ERK and Akt activity dynamics are predictive of stochastic cell division events.

Understanding the dynamics of intracellular signaling pathways, such as ERK1/2 (ERK) and Akt1/2 (Akt), in the context of cell fate decisions is important for advancing our knowledge of cellular processes and diseases, particularly cancer. While previous studies have established associations between ERK and Akt activities and proliferative cell fate, the heterogeneity of single-cell responses adds complexity to this understanding. This study employed a data-driven approach to address this challenge, developing machine learning models trained on a dataset of growth factor-induced ERK and Akt activity time courses in single cells, to predict cell division events. The most predictive models were developed by applying discrete wavelet transforms (DWTs) to extract low-frequency features from the time courses, followed by using Ensemble Integration, a data integration and predictive modeling framework. The results demonstrated that these models effectively predicted cell division events in MCF10A cells (F-measure=0.524, AUC=0.726). ERK dynamics were found to be more predictive than Akt, but the combination of both measurements further enhanced predictive performance. The ERK model`s performance also generalized to predicting division events in RPE cells, indicating the potential applicability of these models and our data-driven methodology for predicting cell division across different biological contexts. Interpretation of these models suggested that ERK dynamics throughout the cell cycle, rather than immediately after growth factor stimulation, were associated with the likelihood of cell division. Overall, this work contributes insights into the predictive power of intra-cellular signaling dynamics for cell fate decisions, and highlights the potential of machine learning approaches in unraveling complex cellular behaviors.

Increasing Signal Intensity of Fluorescent Oligo-Labeled Antibodies to Enable Combination Multiplexing.

Full-spectrum flow cytometry has increased antibody-based multiplexing, yet further increases remain potentially impactful. We recently proposed how fluorescence multiplexing using spectral imaging and combinatorics (MuSIC) could do so using tandem dyes and an oligo-based antibody labeling method. In this work, we found that such labeled antibodies had significantly lower signal intensities than conventionally labeled antibodies in human cell experiments. To improve signal intensity, we tested moving the fluorophores from the original external (ext.) 5' or 3' end-labeled orientation to internal (int.) fluorophore modifications. Cell-free spectrophotometer measurements showed a ∼6-fold signal intensity increase of the new int. configuration compared to the previous ext. configuration. Time-resolved fluorescence and fluorescence correlation spectroscopy showed that the ∼3-fold brightness difference is due to static quenching most likely by the oligo or solution in the ext. configuration. Spectral flow cytometry experiments using peripheral blood mononuclear cells show int. MuSIC probe-labeled antibodies (i) retained increased signal intensity while having no significant difference in the estimated % of CD8+ lymphocytes and (ii) labeled with Atto488, Atto647, and Atto488/647 combinations can be demultiplexed in triple-stained samples. The antibody labeling approach is general and can be broadly applied to many biological and diagnostic applications where spectral detection is available.

A 96-Well Polyacrylamide Gel for Electrophoresis and Western Blotting.

Western blotting is a stalwart technique for analyzing specific proteins and/or their post-translational modifications. However, it remains challenging to accommodate more than ∼10 samples per experiment without substantial departure from trusted, established protocols involving accessible instrumentation. Here, we describe a 96-sample western blot that conforms to standard 96-well plate dimensional constraints and has little operational deviation from standard western blotting. The main differences are that (i) submerged polyacrylamide gel electrophoresis is operated horizontally (similar to agarose gels) as opposed to vertically, and (ii) a 6 mm thick gel is used, with 2 mm most relevant for membrane transfer (vs ∼1 mm typical). Results demonstrate both wet and semi-dry transfer are compatible with this gel thickness. The major tradeoff is reduced molecular weight resolution, due primarily to less available migration distance per sample. We demonstrate proof-of-principle using gels loaded with molecular weight ladder, recombinant protein, and cell lysates. We expect the 96-well western blot will increase reproducibility, efficiency, and capacity for biological characterization relative to established western blots.

Network Analyses of Brain Tumor Patients' Multiomic Data Reveals Pharmacological Opportunities to Alter Cell State Transitions.

Glioblastoma Multiforme (GBM) remains a particularly difficult cancer to treat, and survival outcomes remain poor. In addition to the lack of dedicated drug discovery programs for GBM, extensive intratumor heterogeneity and epigenetic plasticity related to cell-state transitions are major roadblocks to successful drug therapy in GBM. To study these phenomenon, publicly available snRNAseq and bulk RNAseq data from patient samples were used to categorize cells from patients into four cell states (i.e. phenotypes), namely: (i) neural progenitor-like (NPC-like), (ii) oligodendrocyte progenitor-like (OPC-like), (iii) astrocyte-like (AC-like), and (iv) mesenchymal-like (MES-like). Patients were subsequently grouped into subpopulations based on which cell-state was the most dominant in their respective tumor. By incorporating phosphoproteomic measurements from the same patients, a protein-protein interaction network (PPIN) was constructed for each cell state. These four-cell state PPINs were pooled to form a single Boolean network that was used for in silico protein knockout simulations to investigate mechanisms that either promote or prevent cell state transitions. Simulation results were input into a boosted tree machine learning model which predicted the cell states or phenotypes of GBM patients from an independent public data source, the Glioma Longitudinal Analysis (GLASS) Consortium. Combining the simulation results and the machine learning predictions, we generated hypotheses for clinically relevant causal mechanisms of cell state transitions. For example, the transcription factor TFAP2A can be seen to promote a transition from the NPC-like to the MES-like state. Such protein nodes and the associated signaling pathways provide potential drug targets that can be further tested in vitro and support cell state-directed (CSD) therapy.

Theory for High-Throughput Genetic Interaction Screening.

Systematic, genome-scale genetic screens have been instrumental for elucidating genotype-phenotype relationships, but approaches for probing genetic interactions have been limited to at most ∼100 pre-selected gene combinations in mammalian cells. Here, we introduce a theory for high-throughput genetic interaction screens. The theory extends our recently developed Multiplexing using Spectral Imaging and Combinatorics (MuSIC) approach to propose ∼105 spectrally unique, genetically encoded MuSIC barcodes from 18 currently available fluorescent proteins. Simulation studies based on constraints imposed by spectral flow cytometry equipment suggest that genetic interaction screens at the human genome-scale may be possible if MuSIC barcodes can be paired to guide RNAs. While experimental testing of this theory awaits, it offers transformative potential for genetic perturbation technology and knowledge of genetic function. More broadly, the availability of a genome-scale spectral barcode library for non-destructive identification of single cells could find more widespread applications such as traditional genetic screening and high-dimensional lineage tracing.

Increasing Signal Intensity of Fluorescent Oligo-Labeled Antibodies.

While full-spectrum flow cytometry has increased antibody-based multiplexing, yet further increases remain potentially impactful. We recently proposed how fluorescence Multiplexing using Spectral Imaging and Combinatorics (MuSIC) could do so using tandem dyes and an oligo-based antibody labeling method. In this work, we found that such labeled antibodies had significantly lower signal intensity than conventionally-labeled antibodies in human cell experiments. To improve signal intensity, we tested moving the fluorophores from the original external (ext.) 5' or 3' end-labeled orientation to internal (int.) fluorophore modifications. Cell-free spectrophotometer measurements showed a ~6-fold signal intensity increase of the new int. configuration compared to the previous ext. configuration. Time-resolved fluorescence spectroscopy and fluorescence correlation spectroscopy showed that ~3-fold brightness difference is due to static quenching. Spectral flow cytometry experiments using peripheral blood mononuclear cells stained with anti-CD8 antibodies showed that int. MuSIC probe-labeled antibodies have signal intensity equal to or greater than conventionally-labeled antibodies with similar estimated proportion of CD8+ lymphocytes. The antibody labeling approach is general and can be broadly applied to many biological and diagnostic applications.

Modeling the Dynamics of Eukaryotic DNA Synthesis in Remembrance of Tunde Ogunnaike.

Two things Tunde loved were dynamics and probability. The work described herein combined them both, which explains why Tunde invariably asked me each time we talked how this work was proceeding. However, as I've come to appreciate as reminiscent of a surprisingly large amount of work in almost any researcher's career, I did not complete a peer-reviewed article on the matter while he could see it. We were broadly motivated by analysis of data for total DNA content in single cells, across thousands of cells. From such data one can estimate the proportions of cells in different phases of the cell cycle by fitting a mixture model for subpopulations of G0/G1 phase cells (1 relative copy of the genome), S phase cells (between 1 and 2 relative copies of the genome), and G2/M phase cells (2 relative copies of the genome). Given an asynchronously cycling population, Gaussian models are reasonable for the G0/G1 and G2/M subpopulations, but an appropriate functional form for the S-phase subpopulation was unclear. Since the probability of observing an S-phase cell is intimately related to the dynamics of DNA replication, we worked to derive a model for DNA replication dynamics from first principles, resulting in a closed-form, analytic expression for the dynamics of DNA synthesis. While quite arguably a somewhat superfluous effort, there is a certain satisfaction and academic beauty to modeling systems from a first-principles approach, and it can sometimes lead to unexpected scientific insights. Yet, while mathematically elegant, there was a fundamental issue with a key assumption that the so-called inter-origin distance distribution (distances between DNA replication initiation sites) was time-invariant. First, I present the model as developed previously. Then, to address the time-invariant inter-origin distance distribution issue, I provide a treatment of time-varying inter-origin distance distributions that, while mathematically simple, provides (i) mechanistic predictions for how all the DNA in a fertilized frog egg can be replicated "on time" despite some inter-origin distances initially exceeding the corresponding amount of allowable time and (ii) evidence that, based only on data from DNA content versus time and average inter-origin distances, somatic cell DNA is parsed into distinct regions whose replication is temporally separate.

MOBILE pipeline enables identification of context-specific networks and regulatory mechanisms.

Robust identification of context-specific network features that control cellular phenotypes remains a challenge. We here introduce MOBILE (Multi-Omics Binary Integration via Lasso Ensembles) to nominate molecular features associated with cellular phenotypes and pathways. First, we use MOBILE to nominate mechanisms of interferon-γ (IFNγ) regulated PD-L1 expression. Our analyses suggest that IFNγ-controlled PD-L1 expression involves BST2, CLIC2, FAM83D, ACSL5, and HIST2H2AA3 genes, which were supported by prior literature. We also compare networks activated by related family members transforming growth factor-beta 1 (TGFß1) and bone morphogenetic protein 2 (BMP2) and find that differences in ligand-induced changes in cell size and clustering properties are related to differences in laminin/collagen pathway activity. Finally, we demonstrate the broad applicability and adaptability of MOBILE by analyzing publicly available molecular datasets to investigate breast cancer subtype specific networks. Given the ever-growing availability of multi-omics datasets, we envision that MOBILE will be broadly useful for identification of context-specific molecular features and pathways.

Predicting anti-cancer drug combination responses with a temporal cell state network model.

Cancer chemotherapy combines multiple drugs, but predicting the effects of drug combinations on cancer cell proliferation remains challenging, even for simple in vitro systems. We hypothesized that by combining knowledge of single drug dose responses and cell state transition network dynamics, we could predict how a population of cancer cells will respond to drug combinations. We tested this hypothesis here using three targeted inhibitors of different cell cycle states in two different cell lines in vitro. We formulated a Markov model to capture temporal cell state transitions between different cell cycle phases, with single drug data constraining how drug doses affect transition rates. This model was able to predict the landscape of all three different pairwise drug combinations across all dose ranges for both cell lines with no additional data. While further application to different cell lines, more drugs, additional cell state networks, and more complex co-culture or in vivo systems remain, this work demonstrates how currently available or attainable information could be sufficient for prediction of drug combination response for single cell lines in vitro.

Computational speed-up of large-scale, single-cell model simulations via a fully integrated SBML-based format.

Summary: Large-scale and whole-cell modeling has multiple challenges, including scalable model building and module communication bottlenecks (e.g. between metabolism, gene expression, signaling, etc.). We previously developed an open-source, scalable format for a large-scale mechanistic model of proliferation and death signaling dynamics, but communication bottlenecks between gene expression and protein biochemistry modules remained. Here, we developed two solutions to communication bottlenecks that speed-up simulation by ∼4-fold for hybrid stochastic-deterministic simulations and by over 100-fold for fully deterministic simulations. Fully deterministic speed-up facilitates model initialization, parameter estimation and sensitivity analysis tasks. Availability and implementation: Source code is freely available at https://github.com/birtwistlelab/SPARCED/releases/tag/v1.3.0 implemented in python, and supported on Linux, Windows and MacOS (via Docker).

Predicting individual-specific cardiotoxicity responses induced by tyrosine kinase inhibitors.

Introduction: Tyrosine kinase inhibitor drugs (TKIs) are highly effective cancer drugs, yet many TKIs are associated with various forms of cardiotoxicity. The mechanisms underlying these drug-induced adverse events remain poorly understood. We studied mechanisms of TKI-induced cardiotoxicity by integrating several complementary approaches, including comprehensive transcriptomics, mechanistic mathematical modeling, and physiological assays in cultured human cardiac myocytes. Methods: Induced pluripotent stem cells (iPSCs) from two healthy donors were differentiated into cardiac myocytes (iPSC-CMs), and cells were treated with a panel of 26 FDA-approved TKIs. Drug-induced changes in gene expression were quantified using mRNA-seq, changes in gene expression were integrated into a mechanistic mathematical model of electrophysiology and contraction, and simulation results were used to predict physiological outcomes. Results: Experimental recordings of action potentials, intracellular calcium, and contraction in iPSC-CMs demonstrated that modeling predictions were accurate, with 81% of modeling predictions across the two cell lines confirmed experimentally. Surprisingly, simulations of how TKI-treated iPSC-CMs would respond to an additional arrhythmogenic insult, namely, hypokalemia, predicted dramatic differences between cell lines in how drugs affected arrhythmia susceptibility, and these predictions were confirmed experimentally. Computational analysis revealed that differences between cell lines in the upregulation or downregulation of particular ion channels could explain how TKI-treated cells responded differently to hypokalemia. Discussion: Overall, the study identifies transcriptional mechanisms underlying cardiotoxicity caused by TKIs, and illustrates a novel approach for integrating transcriptomics with mechanistic mathematical models to generate experimentally testable, individual-specific predictions of adverse event risk.

MEMMAL: A tool for expanding large-scale mechanistic models with machine learned associations and big datasets.

Computational models that can explain and predict complex sub-cellular, cellular, and tissue-level drug response mechanisms could speed drug discovery and prioritize patient-specific treatments (i.e., precision medicine). Some models are mechanistic with detailed equations describing known (or supposed) physicochemical processes, while some are statistical or machine learning-based approaches, that explain datasets but have no mechanistic or causal guarantees. These two types of modeling are rarely combined, missing the opportunity to explore possibly causal but data-driven new knowledge while explaining what is already known. Here, we explore combining machine learned associations with mechanistic models to develop computational models that could more fully represent cellular behavior. In this proposed MEMMAL (MEchanistic Modeling with MAchine Learning) framework, machine learning/statistical models built using omics datasets provide predictions for new interactions between genes and proteins where there is physicochemical uncertainty. These interactions are used as a basis for new reactions in mechanistic models. As a test case, we focused on incorporating novel IFNγ/PD-L1 related associations into a large-scale mechanistic model for cell proliferation and death to better recapitulate the recently released NIH LINCS Consortium MCF10A dataset and enable description of the cellular response to checkpoint inhibitor immunotherapies. This work is a template for combining big-data-inferred interactions with mechanistic models, which could be more broadly applicable for building multi-scale precision medicine and whole cell models.

Network inference from perturbation time course data.

Networks underlie much of biology from subcellular to ecological scales. Yet, understanding what experimental data are needed and how to use them for unambiguously identifying the structure of even small networks remains a broad challenge. Here, we integrate a dynamic least squares framework into established modular response analysis (DL-MRA), that specifies sufficient experimental perturbation time course data to robustly infer arbitrary two and three node networks. DL-MRA considers important network properties that current methods often struggle to capture: (i) edge sign and directionality; (ii) cycles with feedback or feedforward loops including self-regulation; (iii) dynamic network behavior; (iv) edges external to the network; and (v) robust performance with experimental noise. We evaluate the performance of and the extent to which the approach applies to cell state transition networks, intracellular signaling networks, and gene regulatory networks. Although signaling networks are often an application of network reconstruction methods, the results suggest that only under quite restricted conditions can they be robustly inferred. For gene regulatory networks, the results suggest that incomplete knockdown is often more informative than full knockout perturbation, which may change experimental strategies for gene regulatory network reconstruction. Overall, the results give a rational basis to experimental data requirements for network reconstruction and can be applied to any such problem where perturbation time course experiments are possible.

Relating individual cell division events to single-cell ERK and Akt activity time courses.

Biochemical correlates of stochastic single-cell fates have been elusive, even for the well-studied mammalian cell cycle. We monitored single-cell dynamics of the ERK and Akt pathways, critical cell cycle progression hubs and anti-cancer drug targets, and paired them to division events in the same single cells using the non-transformed MCF10A epithelial line. Following growth factor treatment, in cells that divide both ERK and Akt activities are significantly higher within the S-G2 time window (~ 8.5-40 h). Such differences were much smaller in the pre-S-phase, restriction point window which is traditionally associated with ERK and Akt activity dependence, suggesting unappreciated roles for ERK and Akt in S through G2. Simple metrics of central tendency in this time window are associated with subsequent cell division fates. ERK activity was more strongly associated with division fates than Akt activity, suggesting Akt activity dynamics may contribute less to the decision driving cell division in this context. We also find that ERK and Akt activities are less correlated with each other in cells that divide. Network reconstruction experiments demonstrated that this correlation behavior was likely not due to crosstalk, as ERK and Akt do not interact in this context, in contrast to other transformed cell types. Overall, our findings support roles for ERK and Akt activity throughout the cell cycle as opposed to just before the restriction point, and suggest ERK activity dynamics may be more important than Akt activity dynamics for driving cell division in this non-transformed context.

A multi-omic analysis of MCF10A cells provides a resource for integrative assessment of ligand-mediated molecular and phenotypic responses.

The phenotype of a cell and its underlying molecular state is strongly influenced by extracellular signals, including growth factors, hormones, and extracellular matrix proteins. While these signals are normally tightly controlled, their dysregulation leads to phenotypic and molecular states associated with diverse diseases. To develop a detailed understanding of the linkage between molecular and phenotypic changes, we generated a comprehensive dataset that catalogs the transcriptional, proteomic, epigenomic and phenotypic responses of MCF10A mammary epithelial cells after exposure to the ligands EGF, HGF, OSM, IFNG, TGFB and BMP2. Systematic assessment of the molecular and cellular phenotypes induced by these ligands comprise the LINCS Microenvironment (ME) perturbation dataset, which has been curated and made publicly available for community-wide analysis and development of novel computational methods ( synapse.org/LINCS_MCF10A ). In illustrative analyses, we demonstrate how this dataset can be used to discover functionally related molecular features linked to specific cellular phenotypes. Beyond these analyses, this dataset will serve as a resource for the broader scientific community to mine for biological insights, to compare signals carried across distinct molecular modalities, and to develop new computational methods for integrative data analysis.

Mesowestern Blot: Simultaneous Analysis of Hundreds of Submicroliter Lysates.

Western blotting is a widely used technique for molecular-weight-resolved analysis of proteins and their posttranslational modifications, but high-throughput implementations of the standard slab gel arrangement are scarce. The previously developed Microwestern requires a piezoelectric pipetting instrument, which is not available for many labs. Here, we report the Mesowestern blot, which uses a 3D-printable gel casting mold to enable high-throughput Western blotting without piezoelectric pipetting and is compatible with the standard sample preparation and small (∼1 µL) sample sizes. The main tradeoffs are reduced molecular weight resolution and higher sample-to-sample CV, making it suitable for qualitative screening applications. The casted polyacrylamide gel contains 336, ∼0.5 µL micropipette-loadable sample wells arranged within a standard microplate footprint. Polyacrylamide % can be altered to change molecular weight resolution profiles. Proof-of-concept experiments using both infrared-fluorescent molecular weight protein ladder and cell lysate (RIPA buffer) demonstrate that the protein loaded in Mesowestern gels is amenable to the standard Western blotting steps. The main difference between Mesowestern and traditional Western is that semidry horizontal instead of immersed vertical gel electrophoresis is used. The linear range of detection is at least 32-fold, and at least ∼500 attomols of ß-actin can be detected (∼29 ng of total protein from mammalian cell lysates: ∼100-300 cells). Because the gel mold is 3D-printable, users with access to additive manufacturing cores have significant design freedom for custom layouts. We expect that the technique could be easily adopted by any typical cell and molecular biology laboratory already performing Western blots.

Method for Improved Fluorescence Corrections for Molar Mass Characterization by Multiangle Light Scattering.

Multiangle light scattering (MALS) was used to determine the absolute molar mass of fluorescent macromolecules. It is standard protocol to install bandwidth filters before MALS detectors to suppress detection of fluorescent emissions. Fluorescence can introduce tremendous error in light scattering measurements and is a formidable challenge in accurately characterizing fluorescent macromolecules and particles. However, we show that for some systems, bandwidth filters alone are insufficient for blocking fluorescence in molar mass determinations. For these systems, we have devised a correction procedure to calculate the amount of fluorescence interference in the filtered signal. By determining the intensity of fluorescent emission not blocked by the bandwidth filters, we can correct the filtered signal accordingly and accurately determine the true molar mass. The transmission rates are calculated before MALS experimentation using emission data from standard fluorimetry techniques, allowing for the characterization of unknown samples. To validate the correction procedure, we synthesized fluorescent dye-conjugated proteins using an IR800CW (LI-COR) fluorophore and Bovine Serum Albumin protein. We successfully eliminated fluorescence interference in MALS measurements using this approach. This correction procedure has potential application toward more accurate molar mass characterizations of macromolecules with intrinsic fluorescence, such as lignins, fluorescent proteins, fluorescence-tagged proteins, and optically active nanoparticles.

A scalable, open-source implementation of a large-scale mechanistic model for single cell proliferation and death signaling.

Mechanistic models of how single cells respond to different perturbations can help integrate disparate big data sets or predict response to varied drug combinations. However, the construction and simulation of such models have proved challenging. Here, we developed a python-based model creation and simulation pipeline that converts a few structured text files into an SBML standard and is high-performance- and cloud-computing ready. We applied this pipeline to our large-scale, mechanistic pan-cancer signaling model (named SPARCED) and demonstrate it by adding an IFNγ pathway submodel. We then investigated whether a putative crosstalk mechanism could be consistent with experimental observations from the LINCS MCF10A Data Cube that IFNγ acts as an anti-proliferative factor. The analyses suggested this observation can be explained by IFNγ-induced SOCS1 sequestering activated EGF receptors. This work forms a foundational recipe for increased mechanistic model-based data integration on a single-cell level, an important building block for clinically-predictive mechanistic models.

Leveraging modeling and simulation to optimize the therapeutic window for epigenetic modifier drugs.

Dysregulated epigenetic processes can lead to altered gene expression and give rise to malignant transformation and tumorigenesis. Epigenetic drugs aim to revert the phenotype of cancer cells to normally functioning cells, and are developed and applied to treat both hematological and solid cancers. Despite this promising therapeutic avenue, the successful development of epigenetic modulators has been challenging. We argue that besides identifying the right responder patient population, the selection of an optimized dosing regimen is equally important. For the majority of epigenetic modulators, hematological adverse effects such as thrombocytopenia, anemia or neutropenia are frequently observed and may limit their therapeutic potential. Therefore, one of the key challenges is to identify a dosing regimen that maximizes drug efficacy and minimizes toxicity. This requires a good understanding of the quantitative relationship between the administered dose, the drug exposure and the magnitude and duration of drug response related to safety and efficacy. With case examples, we highlight how modeling and simulation has been successfully applied to address those questions. As an outlook, we suggest the combination of efficacy and safety prediction models that capture the quantitative, mechanistic relationships governing the balance between their safety and efficacy dynamics. A stepwise approach for its implementation is presented. Utilizing in silico explorations, the impact of dosing regimen on the therapeutic window can be explored. This will serve as a basis to select the most promising dosing regimen that maximizes efficacy while minimizing adverse effects and to increase the probability of success for the given epigenetic drug.