Background
SANGHA sits at the intersection of two trajectories: quantitative translational science and research-intelligence practice. The first establishes technical depth; the second makes explicit how heterogeneous evidence is turned into bounded scientific decisions.
Scientific background
PhD-trained computational biologist and biostatistician with roughly a decade of work spanning neurological biomarker discovery, metabolomics, statistical genomics, neurodegeneration, postmortem MRI, and reproducible scientific software.
The technical surface includes multivariate modeling, regularized regression, Bayesian and co-expression networks, GWAS summary statistics, PRS decomposition, allele harmonization, compositional analysis, public-reference genomics, Python/R, workflow engineering, provenance, and HPC-aware execution.
Research practice
Across domains, the recurring operation has been the same: acquire heterogeneous evidence, choose a representation, preserve missingness and uncertainty, test sensitivity, distinguish what is established from what is merely suggestive, and identify the next observation that would change the decision.
SANGHA is the effort to make that operation explicit enough to engineer: inspectable scientific objects rather than private judgment alone.
Plasma metabolomics with clinical collaborators; phenotype-level question prioritization and multivariate latent-structure methods.
Targeted and nontargeted metabolomics, biomarker panels, mixed models, network analysis, pathway interpretation, and cross-cohort neurodegeneration work.
Established new postmortem MRI capability, coordinated imaging and engineering requirements, and supported multimodal spatial co-registration.
Longitudinal prediabetes biomarker analysis within a large multi-institution program.
Statistical genomics, phenotype geometry, inferential audit, scientific software, and Theoretical & Computational Medicine under SANGHA Research Systems.