About

Daniel Naiman is a professor of applied mathematics and statistics, serving as the associate department head and director of graduate studies at the Department of Applied Mathematics and Statistics at Johns Hopkins University. He specializes in statistics, computational probability, and bioinformatics, developing methods and tools to analyze complex biological data. His research encompasses various areas of statistical application, including genetics, bioinformatics, and environmental data analysis related to human health risk assessment. Supported by the National Science Foundation, Naiman has developed a series of inferential statistical methods and geometric tools to advance computational statistical theory and methodology. His early research focused on the intersection of geometry, computing, and applied statistics, particularly on the problem of p-value computation in multiple testing. This work led to the development of methodology-extending tools in differential geometry, such as the Hotelling-Weyl tube formula, and the co-development of a theory of abstract tubes, which exploits geometric structures in statistical models to improve inclusion-exclusion identities and importance sampling methods. These contributions culminated in the formulation known as Naiman’s Inequality. Naiman has also contributed to STEM education, creating a summer course called Mathematics of Music to attract students to mathematics. He has received numerous teaching awards, including the 2019 William H. Huggins Excellence in Teaching Award, and has authored over 70 publications and two books. Naiman earned his AB in Mathematics from Cornell University, and his MS in Statistics and PhD in Mathematics from the University of Illinois at Urbana-Champaign.

Research topics

• Medicine
• Environmental health
• Chemistry
• Computer Science
• Biology
• Environmental chemistry
• Physical therapy
• Psychology
• Mathematics
• Internal medicine
• Pediatrics
• Physiology
• Animal science
• Endocrinology
• Nursing

Selected publications

• Sharpened localization of the trailing point of the Pareto record frontier

  Theoretical Computer Science · 2026-02-18

  article
  PublisherDOI

• CellCover Defines Marker Gene Panels Capturing Developmental Progression in Neocortical Neural Stem Cell Identity

  eLife · 2025-10-21

  preprintOpen access

  Definition of cell classes across the tissues of living organisms is central in the analysis of growing atlases of single-cell RNA sequencing (scRNA-seq) data across biomedicine. Marker genes for cell classes are most often defined by differential expression (DE) methods that serially assess individual genes across landscapes of diverse cells. This serial approach has been extremely useful, but is limited because it ignores possible redundancy or complementarity across genes that can only be captured by analyzing multiple genes simultaneously. Interrogating binarized expression data, we aim to identify discriminating panels of genes that are specific to, not only enriched in, individual cell types. To efficiently explore the vast space of possible marker panels, leverage the large number of cells often sequenced, and overcome zero-inflation in scRNA-seq data, we propose viewing marker gene panel selection as a variation of the “minimal set-covering problem” in combinatorial optimization. Using scRNA-seq data from blood and brain tissue, we show that this new method, CellCover, performs as good or better than DE and other methods in defining cell-type discriminating gene panels, while reducing gene redundancy and capturing cell-class-specific signals that are distinct from those defined by DE methods. Transfer learning experiments across mouse, primate, and human data demonstrate that CellCover identifies markers of conserved cell classes in neocortical neurogenesis, as well as developmental progression in both progenitors and neurons. Exploring markers of human outer radial glia (oRG, or basal RG) across mammals, we show that transcriptomic elements of this key cell type in the expansion of the human cortex likely appeared in gliogenic precursors of the rodent before the full program emerged in neurogenic cells of the primate lineage. We have assembled the public datasets we use in this report within the NeMO Analytics multi-omic data exploration environment [1], where the expression of individual genes (NeMO: Individual genes in cortex and NeMO: Individual genes in blood) and marker gene panels (NeMO: Telley 3 CellCover Panels, NeMO: Telley 12 CellCover Panels, NeMO: Sorted Brain Cell CellCover Panels, and NeMO: Blood 34 CellCover Panels) can be freely explored without coding expertise. CellCover is available in CellCover R and CellCover Python.

  PublisherDOI

• Are NHANES Data Representative of the US Population for Chemicals with Seasonal and Regional Use?

  Environmental Health Perspectives · 2025-04-25 · 1 citations

  articleOpen access
  PublisherOA PDFDOI

• CellCover Defines Marker Gene Panels Capturing Developmental Progression in Neocortical Neural Stem Cell Identity

  eLife · 2025-10-21 · 1 citations

  preprintOpen access

  Definition of cell classes across the tissues of living organisms is central in the analysis of growing atlases of single-cell RNA sequencing (scRNA-seq) data across biomedicine. Marker genes for cell classes are most often defined by differential expression (DE) methods that serially assess individual genes across landscapes of diverse cells. This serial approach has been extremely useful, but is limited because it ignores possible redundancy or complementarity across genes that can only be captured by analyzing multiple genes simultaneously. Interrogating binarized expression data, we aim to identify discriminating panels of genes that are specific to, not only enriched in, individual cell types. To efficiently explore the vast space of possible marker panels, leverage the large number of cells often sequenced, and overcome zero-inflation in scRNA-seq data, we propose viewing marker gene panel selection as a variation of the “minimal set-covering problem” in combinatorial optimization. Using scRNA-seq data from blood and brain tissue, we show that this new method, CellCover, performs as good or better than DE and other methods in defining cell-type discriminating gene panels, while reducing gene redundancy and capturing cell-class-specific signals that are distinct from those defined by DE methods. Transfer learning experiments across mouse, primate, and human data demonstrate that CellCover identifies markers of conserved cell classes in neocortical neurogenesis, as well as developmental progression in both progenitors and neurons. Exploring markers of human outer radial glia (oRG, or basal RG) across mammals, we show that transcriptomic elements of this key cell type in the expansion of the human cortex likely appeared in gliogenic precursors of the rodent before the full program emerged in neurogenic cells of the primate lineage. We have assembled the public datasets we use in this report within the NeMO Analytics multi-omic data exploration environment [1], where the expression of individual genes (NeMO: Individual genes in cortex and NeMO: Individual genes in blood) and marker gene panels (NeMO: Telley 3 CellCover Panels, NeMO: Telley 12 CellCover Panels, NeMO: Sorted Brain Cell CellCover Panels, and NeMO: Blood 34 CellCover Panels) can be freely explored without coding expertise. CellCover is available in CellCover R and CellCover Python.

  PublisherDOI

• Sharpened Localization of the Trailing Point of the Pareto Record Frontier

  arXiv (Cornell University) · 2024-01-01

  preprintOpen access

  For $d\ge2$ and iid $d$-dimensional observations $X^{(1)},X^{(2)},\dots$ with independent Exponential$(1)$ coordinates, we revisit the study by Fill and Naiman (Electron. J. Probab., 2020) of the boundary (relative to the closed positive orthant), or "frontier", $F_n$ of the closed Pareto record-setting (RS) region \[ \mbox{RS}_n:=\{0\le x\in{\mathbb R}^d:x\not\prec X^{(i)}\mbox{\ for all $1\le i\le n$}\} \] at time $n$, where $0\le x$ means that $0\le x_j$ for $1\le j\le d$ and $x\prec y$ means that $x_j0$ and $c_n\to\infty$ we have \[ {\mathbb P}(F_n^- -\ln n\in (-(2+\varepsilon)\ln\ln\ln n,c_n))\to 1 \] (describing typical behavior) and almost surely \[ \limsup \frac{F_n^- - \ln n}{\ln \ln n} \le 0 \quad \mbox{and} \quad \liminf \frac{F_n^- - \ln n}{\ln \ln \ln n} \in [-2, -1]. \] In this paper we use the theory of generators (minima of $F_n$) together with the first- and second-moment methods to improve considerably the trailing-point location results to \[ F_n^- - (\ln n - \ln \ln \ln n) \overset{\mathrm{P}}{\longrightarrow} - \ln(d - 1) \] (describing typical behavior) and, for $d \ge 3$, almost surely \begin{align*} &\limsup [F_n^- - (\ln n - \ln \ln \ln n)] \leq -\ln(d - 2) + \ln 2 \\ \mbox{and }&\liminf [F_n^- - (\ln n - \ln \ln \ln n)] \ge - \ln d - \ln 2. \end{align*}

  PublisherOA PDFDOI

• Elusive boundaries: using an attribute framework to describe systems for population physical activity promotion

  Health Promotion International · 2024-02-01 · 3 citations

  articleOpen access

  The cost of physical inactivity is alarming, and calls for whole-of-system approaches to population physical activity promotion (PPAP) are increasing. One innovative approach to PPAP is to use a framework of interdependent attributes and associated dimensions of effective systems for chronic disease prevention. Describing system boundaries can be an elusive task, and this article reports on using an attribute framework as a first step in describing and then assessing and strengthening a provincial system for PPAP in British Columbia, Canada. Interviews were conducted with provincial stakeholders to gather perspectives regarding attributes of the system. Following this, two workshops were facilitated to document important stories about the current system for PPAP and link story themes with attributes. Results from interviews and workshops were summarized into key findings and a set of descriptive statements. One hundred and twenty-one statements provide depth, breadth and scope to descriptions of the system through the lens of an adapted framework including four attributes: (i) implementation of desired actions, (ii) resources, (iii) leadership and (iv) collaborative capacity. The attribute framework was a useful tool to guide a whole-of-system approach and turn elusive boundaries into rich descriptors of a provincial system for PPAP. Immediate implications for our research are to translate descriptive statements into variables, then assess the system through group model building and identify leverage points from a causal loop diagram to strengthen the system. Future application of this approach in other contexts, settings and health promotion and disease prevention topics is recommended.

  PublisherDOI

• Follow the Arrows: Using a Co-Created Causal Loop Diagram to Explore Leverage Points to Strengthen Population Physical Activity Promotion in British Columbia, Canada

  Journal of Physical Activity and Health · 2024-05-10 · 2 citations

  articleOpen access

  BACKGROUND: Population physical activity promotion (PPAP) is one of the most effective noncommunicable disease prevention strategies, yet coordination is lacking around the world. Whole-of-system approaches and complex systems methods are called for to advance PPAP. This paper reports on a project which (1) used an Attributes Framework with system mapping (group model building and causal loop diagramming of feedback loops) and (2) identified potential leverage points to address the challenge of effective coordination of multisectoral PPAP in British Columbia. METHODS: Key findings from stakeholder interviews and workshops described the current system for PPAP in terms of attributes and dimensions in the framework. These were translated into variables and used in group model building. Participants prioritized the importance of variables to address the coordination challenge and then created causal loop diagrams in 3 small groups. One collective causal loop diagram was created, and top priority variables and associated feedback loops were highlighted to explore potential leverage points. RESULTS: Leverage points included the relationships and feedback loops among priority variables: political leadership, visible policy support and governance, connectivity for knowledge translation, collaborative multisector grants, multisector collaboration, and integrating co-benefits. Leveraging and altering "vicious" cyclical patterns to increase coordinated multisector PPAP are key. CONCLUSIONS: The Attributes Framework, group model building and causal loop diagrams, and emergent feedback loops were useful to explore potential leverage points to address the challenge of multisectoral coordination of PPAP. Future research could apply the same methods in other jurisdictions and compare and contrast resultant frameworks, variables, feedback loops, and leverage points.

  PublisherOA PDFDOI

• CellCover Defines Marker Gene Panels Capturing Developmental Progression in Neocortical Neural Stem Cell Identity

  bioRxiv (Cold Spring Harbor Laboratory) · 2023-04-07 · 2 citations

  preprintOpen access

  1 Abstract Definition of cell classes across the tissues of living organisms is central in the analysis of growing atlases of single-cell RNA sequencing (scRNA-seq) data across biomedicine. Marker genes for cell classes are most often defined by differential expression (DE) methods that serially assess individual genes across landscapes of diverse cells. This serial approach has been extremely useful, but is limited because it ignores possible redundancy or complementarity across genes that can only be captured by analyzing multiple genes simultaneously. Interrogating binarized expression data, we aim to identify discriminating panels of genes that are specific to, not only enriched in, individual cell types. To efficiently explore the vast space of possible marker panels, leverage the large number of cells often sequenced, and overcome zero-inflation in scRNA-seq data, we propose viewing marker gene panel selection as a variation of the “minimal set-covering problem” in combinatorial optimization. Using scRNA-seq data from blood and brain tissue, we show that this new method, CellCover, performs as good or better than DE and other methods in defining cell-type discriminating gene panels, while reducing gene redundancy and capturing cell-class-specific signals that are distinct from those defined by DE methods. Transfer learning experiments across mouse, primate, and human data demonstrate that CellCover identifies markers of conserved cell classes in neocortical neurogenesis, as well as developmental progression in both progenitors and neurons. Exploring markers of human outer radial glia (oRG, or basal RG) across mammals, we show that transcriptomic elements of this key cell type in the expansion of the human cortex likely appeared in gliogenic precursors of the rodent before the full program emerged in neurogenic cells of the primate lineage. We have assembled the public datasets we use in this report within the NeMO Analytics multi-omic data exploration environment [1], where the expression of individual genes ( NeMO: Individual genes in cortex and NeMO: Individual genes in blood ) and marker gene panels ( NeMO: Telley 3 CellCover Panels , NeMO: Telley 12 CellCover Panels , NeMO: Sorted Brain Cell CellCover Panels , and NeMO: Blood 34 CellCover Panels ) can be freely explored without coding expertise. CellCover is available in CellCover R and CellCover Python . Graphical Abstract

  PublisherOA PDFDOI

• Erratum: Current Breast Milk PFAS Levels in the United States and Canada: After All This Time, Why Don’t We Know More?

  Environmental Health Perspectives · 2023-03-01 · 2 citations

  erratumOpen access

  During the paper editing process, two values in Table 2 were mistakenly modified.Milk:serum concentration ratios from the "Krrman et al. 2007" study should be 0.01 for PFOS and 0.01 PFNA.This error affects values in Table 2 only.All calculations were performed with the correct values.The authors regret the error.

  PublisherDOI

• A scoping review of complex systems methods used in population physical activity research: do they align with attributes of a whole system approach?

  Health Research Policy and Systems · 2023-03-02 · 40 citations

  reviewOpen access

  BACKGROUND: Complex systems approaches are increasingly used in health promotion and noncommunicable disease prevention research, policy and practice. Questions emerge as to the best ways to take a complex systems approach, specifically with respect to population physical activity (PA). Using an Attributes Model is one way to understand complex systems. We aimed to examine the types of complex systems methods used in current PA research and identify what methods align with a whole system approach as reflected by an Attributes Model. METHODS: A scoping review was conducted and two databases were searched. Twenty-five articles were selected and data analysis was based upon the following: the complex systems research methods used, research aims, if participatory methods were used and evidence of discussion regarding attributes of systems. RESULTS: There were three groups of methods used: system mapping, simulation modelling and network analysis. System mapping methods appeared to align best with a whole system approach to PA promotion because they largely aimed to understand complex systems, examined interactions and feedback among variables, and used participatory methods. Most of these articles focused on PA (as opposed to integrated studies). Simulation modelling methods were largely focused on examining complex problems and identifying interventions. These methods did not generally focus on PA or use participatory methods. While network analysis articles focused on examining complex systems and identifying interventions, they did not focus on PA nor use participatory methods. All attributes were discussed in some way in the articles. Attributes were explicitly reported on in terms of findings or were part of discussion and conclusion sections. System mapping methods appear to be well aligned with a whole system approach because these methods addressed all attributes in some way. We did not find this pattern with other methods. CONCLUSIONS: Future research using complex systems methods may benefit from applying the Attributes Model in conjunction with system mapping methods. Simulation modelling and network analysis methods are seen as complementary and could be used when system mapping methods identify priorities for further investigation (e.g. what interventions to implement or how densely connected relationships are in systems).

  PublisherOA PDFDOI

Frequent coauthors

• Judy S. LaKind

  45 shared

• George R. Uhl

  Universität Greifswald

  13 shared

• Qing‐Rong Liu

  12 shared

• Henry P. Wynn

  Indiana University – Purdue University Indianapolis

  12 shared

• Lingxin Hao

  11 shared

• Bret Cooper

  Agricultural Research Service

  11 shared

• Judith Hess

  National Institute on Drug Abuse

  11 shared

• James D. Malley

  Alberta Health Services

  10 shared

Awards & honors

• Fellow of the Institute for Mathematical Statistics (1997)
• Professor Joel Dean Award for Excellence in Teaching (2018)
• Hopkins Undergraduate Society for Applied Mathematics and St…
• William H. Huggins Excellence in Teaching Award (2019)