Ongoing RESEARCH
Ongoing RESEARCH
Embedded Files
PraxisModus Institute for Advanced Studies develops hybrid mechanistic, statistical, and AI/ML modeling frameworks for infectious-disease forecasting, risk assessment, and intervention planning. In parallel, we pursue foundational research in statistical methodology for ecological scaling laws and theoretical ecology of trophic transfer. The following studies illustrate our growing research portfolio across epidemic modeling, climate-sensitive disease prediction, vector-borne disease ecology, policy-oriented decision support, statistical benchmarking, and geometric theories of trophic organization.
PraxisModus Institute for Advanced Studies develops hybrid mechanistic, statistical, and AI/ML modeling frameworks for infectious-disease forecasting, risk assessment, and intervention planning. In parallel, we pursue foundational research in statistical methodology for ecological scaling laws and theoretical ecology of trophic transfer. The following studies illustrate our growing research portfolio across epidemic modeling, climate-sensitive disease prediction, vector-borne disease ecology, policy-oriented decision support, statistical benchmarking, and geometric theories of trophic organization.
Together, the completed and ongoing studies demonstrate the PraxisModus research strategy: combining mechanistic biological understanding, advanced statistical modeling, AI/ML forecasting, intervention simulation, and foundational theoretical ecology to address real-world problems.
Together, the completed and ongoing studies demonstrate the PraxisModus research strategy: combining mechanistic biological understanding, advanced statistical modeling, AI/ML forecasting, intervention simulation, and foundational theoretical ecology to address real-world problems.
These projects form the basis for a broader PraxisModus infectious-disease modeling platform capable of integrating surveillance data, climate data, vector ecology, spatial mobility, and AI-driven forecasting, alongside a parallel effort in theoretical ecology and statistical methodology for scaling laws.
These projects form the basis for a broader PraxisModus infectious-disease modeling platform capable of integrating surveillance data, climate data, vector ecology, spatial mobility, and AI-driven forecasting, alongside a parallel effort in theoretical ecology and statistical methodology for scaling laws.
Geometric Ecology: Trophic Transfer as a Multiplicative Power Law
Geometric Ecology: Trophic Transfer as a Multiplicative Power Law
Geometric Ecology: Trophic Transfer as a Multiplicative Power Law
Geometric Ecology: Trophic Transfer as a Multiplicative Power Law
Geometric Ecology: Trophic Transfer as a Multiplicative Power Law
Geometric Ecology: Trophic Transfer as a Multiplicative Power Law
This ongoing work proposes that affine biomass scaling in logarithmic coordinates imposes a geometric constraint on trophic transfer. Under a broad reduced consumer-resource balance with nonlinear closure, exact affine biomass scaling occurs if and only if trophic transfer is multiplicative and power-law in prey biomass. Thus the power-law response is not merely an empirical curve; under the balance considered, it is the unique exact transfer form compatible with global affine biomass geometry. Classical additive-saturation responses (Holling, Hill, Ivlev) are biologically meaningful but belong to a different geometric class because their additive denominator or exponential structures necessarily introduce curvature in log-space.
The theoretical framework distinguishes the fixed-exponent power-law model (the exact geometry-preserving law) from an extended power-law model (xPL) in which the biomass-dependent exponent $\alpha(B) = 2K^{\xi}/(B^{\xi}+K^{\xi})$ allows controlled departures from constant-exponent scaling while retaining multiplicative logarithmic structure. The analysis proves recursive propagation of affine geometry through finite trophic chains, showing that the same multiplicative structure extends from phytoplankton-zooplankton to higher predator levels.
Simulation-based structural diagnostics, grazing-rate data, and plankton biomass validation across ocean habitats are being used to test whether multiplicative geometric models are favored or competitive across diverse oceanographic regimes. The work establishes a geometry-based framework for trophic model selection and suggests that plankton trophic transfer may be constrained by emergent biomass geometry rather than by any single local saturation mechanism.
The theoretical framework distinguishes the fixed-exponent power-law model (the exact geometry-preserving law) from an extended power-law model (xPL) in which the biomass-dependent exponent $\alpha(B) = 2K^{\xi}/(B^{\xi}+K^{\xi})$ allows controlled departures from constant-exponent scaling while retaining multiplicative logarithmic structure. The analysis proves recursive propagation of affine geometry through finite trophic chains, showing that the same multiplicative structure extends from phytoplankton-zooplankton to higher predator levels.
Simulation-based structural diagnostics, grazing-rate data, and plankton biomass validation across ocean habitats are being used to test whether multiplicative geometric models are favored or competitive across diverse oceanographic regimes. The work establishes a geometry-based framework for trophic model selection and suggests that plankton trophic transfer may be constrained by emergent biomass geometry rather than by any single local saturation mechanism.
Manuscript in preparation; target journal Physical Review E or The American Naturalist
Leishmaniasis: Environmental Drivers of Disease Presence in Sri Lanka
Leishmaniasis: Environmental Drivers of Disease Presence in Sri Lanka
Leishmaniasis: Environmental Drivers of Disease Presence in Sri Lanka
Leishmaniasis: Environmental Drivers of Disease Presence in Sri Lanka
Leishmaniasis: Environmental Drivers of Disease Presence in Sri Lanka
Leishmaniasis: Environmental Drivers of Disease Presence in Sri Lanka
This study examines environmental correlates of cutaneous leishmaniasis presence across Sri Lanka, focusing on climate, soil moisture, and microenvironmental variables relevant to sand fly ecology. Using data from 233 subdistricts, the study applied binary logistic regression to evaluate whether reported leishmaniasis presence was associated with ten standardized environmental predictors. The analysis identified several significant environmental associations. Higher diurnal temperature range, wind speed, solar radiation, and rainfall were associated with increased odds of disease presence, whereas higher surface soil wetness showed a strong protective association. This study represents an early step toward climate-informed leishmaniasis risk mapping in Sri Lanka.
Abstract published in the iPURSE International Annual Research Symposium, University of Peradeniya, Sri Lanka. [Add year, abstract number, and symposium citation when available.]
RAIML
RAIML
RAIML
RAIML
RAIML
RAIML
RAIML is an ongoing open-source software project for immune-cell composition inference from bulk tumor RNA-sequencing data across multiple cancer types. Unlike classical deconvolution tools that treat bulk expression as a linear mixture of predefined immune signatures, RAIML reframes the problem as supervised machine learning: it learns nonlinear mappings from bulk tumor transcriptomes to immune-cell fractions using paired single-cell RNA-seq data as ground truth.
The project aims to convert a validated proof-of-concept into a reusable, documented, API-first software framework for cancer immunology and precision oncology research. RAIML will support multi-cancer benchmarking against established tools such as CIBERSORTx, TIMER, xCell, MCP-counter, and quanTIseq, with the goal of improving immune infiltration estimation, tumor microenvironment analysis, biomarker discovery, and reproducible downstream modeling from existing public datasets.
The project aims to convert a validated proof-of-concept into a reusable, documented, API-first software framework for cancer immunology and precision oncology research. RAIML will support multi-cancer benchmarking against established tools such as CIBERSORTx, TIMER, xCell, MCP-counter, and quanTIseq, with the goal of improving immune infiltration estimation, tumor microenvironment analysis, biomarker discovery, and reproducible downstream modeling from existing public datasets.
TissueInflammMap
TissueInflammMap
TissueInflammMap
TissueInflammMap
TissueInflammMap
TissueInflammMap
TissueInflammMap is an ongoing open-source project developing a hybrid AI–mechanistic framework for mapping tissue-specific inflammation from bulk RNA-seq and related public datasets. The project aims to identify inflammatory gene programs, cytokine and chemokine signatures, immune-cell infiltration patterns, and disease-associated tissue states across diverse human organs and pathological conditions.
By combining curated inflammation biology with machine learning, graph-based modeling, and mechanistic dynamic models, TissueInflammMap is designed to move beyond descriptive gene-expression analysis toward interpretable, prediction-ready inflammation mapping. The platform will support disease-risk stratification, biomarker discovery, therapeutic target prioritization, and in silico simulation of gene, pathway, or drug perturbations for inflammatory disease, cancer, neuroinflammation, metabolic disease, and precision medicine research.
By combining curated inflammation biology with machine learning, graph-based modeling, and mechanistic dynamic models, TissueInflammMap is designed to move beyond descriptive gene-expression analysis toward interpretable, prediction-ready inflammation mapping. The platform will support disease-risk stratification, biomarker discovery, therapeutic target prioritization, and in silico simulation of gene, pathway, or drug perturbations for inflammatory disease, cancer, neuroinflammation, metabolic disease, and precision medicine research.
DHEWS-OS
DHEWS-OS
DHEWS-OS
DHEWS-OS
DHEWS-OS
DHEWS-OS
DHEWS-OS is a submitted open-source software project for a disease early-warning and risk-assessment system, initially developed around Sri Lanka as a national testbed. The platform is designed to integrate weekly infectious-disease surveillance data with climate, environmental, vegetation, population, livestock, and spatial-risk data from public sources to support timely forecasting and risk mapping across districts and subdistricts.
The project aims to provide an adaptable, country-ready framework for multi-disease surveillance, short-term outbreak prediction, hotspot detection, and intervention planning. By combining mechanistic epidemiological models, AI/ML forecasting, satellite-derived environmental data, and interactive dashboards, DHEWS-OS will help public-health teams identify emerging risks earlier and support evidence-based decisions for dengue, leishmaniasis, leptospirosis, anaplasmosis, and other climate-sensitive infectious diseases.
The project aims to provide an adaptable, country-ready framework for multi-disease surveillance, short-term outbreak prediction, hotspot detection, and intervention planning. By combining mechanistic epidemiological models, AI/ML forecasting, satellite-derived environmental data, and interactive dashboards, DHEWS-OS will help public-health teams identify emerging risks earlier and support evidence-based decisions for dengue, leishmaniasis, leptospirosis, anaplasmosis, and other climate-sensitive infectious diseases.
SenesceNet
SenesceNet
SenesceNet
SenesceNet
SenesceNet
SenesceNet
SenesceNet is an ongoing open-source software project that develops a hybrid AI–mechanistic framework for studying tissue-specific aging and aging-related disease risk from bulk RNA-seq data. The project combines graph neural networks, ordinary differential equation modeling, and public resources such as GTEx and TCGA to learn how senescence-associated gene networks change across tissues such as brain, liver, lung, heart, skin, and blood.
The goal is to move beyond black-box aging clocks by building an interpretable platform that can estimate biological aging status, identify tissue-specific senescence programs, predict disease-risk patterns, and simulate counterfactual gene or pathway interventions. SenesceNet is being designed as a reusable, documented, and community-accessible toolkit for aging research, geroscience, biomarker discovery, and precision medicine.
The goal is to move beyond black-box aging clocks by building an interpretable platform that can estimate biological aging status, identify tissue-specific senescence programs, predict disease-risk patterns, and simulate counterfactual gene or pathway interventions. SenesceNet is being designed as a reusable, documented, and community-accessible toolkit for aging research, geroscience, biomarker discovery, and precision medicine.
Page updated
Google Sites
Report abuse