Senior Causal Inference & Mathematical Modeling Scientist (Hybrid Systems | Bayesian Causality | Systems Biology)

Role Summary

Ayass Bioscience is building a next-generation hybrid causal inference platform (BiRAGAS) that integrates established biological knowledge with data-driven discovery to produce interpretable, regulatory-defensible causal models.

We are seeking a senior scientist with deep expertise in mathematical modeling, causal inference, and Bayesian methods to design and implement the mathematical foundations of our hybrid architecture. This role complements our existing machine learning and bioinformatics team by owning the formal equations, priors, and inference machinery that translate biology into rigorous causal models.

This is not a standard ML role. It is a foundational modeling role at the core of the platform.

Core Responsibilities

1. Mathematical Formulation of the Hybrid Causal Framework

• Translate biological knowledge graphs (pathways, directional mechanisms, interventions) into formal mathematical representations
• Define and implement Bayesian priors over causal graph structures (e.g., ( P(G) \propto \exp(\sum \theta_{ij}) ))
• Formalize constraints, forbidden edges, soft priors, and fixed anchors within causal discovery algorithms
• Ensure mathematical consistency across graph structure learning, parameter estimation, and uncertainty quantification

2. Causal Discovery Under Biological Constraints

• Design and adapt constrained causal discovery algorithms (PC, GES, score-based, hybrid methods)
• Incorporate biological directionality, pathway topology, genetic anchors (eQTL/pQTL), and intervention data as first-class constraints
• Address known causal challenges:
• Markov equivalence
• Hidden confounding
• Finite sample limitations
• High-dimensional gene expression spaces

3. Structural Equation & Effect Size Modeling

• Develop and fit Structural Equation Models (SEMs) on biologically constrained graphs
• Estimate context-specific causal effect sizes with confidence intervals
• Support heterogeneous effects, moderators, and disease- or tissue-specific contexts

4. Evidence Integration & Causal Confidence Scoring

• Define the mathematical framework for integrating statistical evidence, priors, genetic evidence, and mechanistic plausibility
• Contribute to a composite causal confidence score that moves beyond p-values toward actionable inference
• Design principled approaches to resolve conflicts between data-driven signals and database knowledge

5. Cross-Functional Collaboration

• Work closely with:
• ML engineers (who implement scalable systems)
• Bioinformaticians (who prepare and interpret omics data)
• Domain scientists (who curate biological knowledge)
• Act as the mathematical authority bridging biology and machine learning

Required Expertise

Mathematical & Statistical Background

• PhD (or equivalent depth) in Applied Mathematics, Statistics, Physics, Computer Science, or related field
• Deep expertise in:
• Bayesian inference
• Probabilistic graphical models
• Causal inference theory
• Optimization and likelihood-based modeling

Causal Inference & Modeling

• Hands-on experience with:
• DAGs, SCMs, SEMs
• Score-based and constraint-based causal discovery
• Priors over graph structures
• Confounding and identifiability
• Strong understanding of why purely data-driven causality fails in biological systems

Computational Skills

• Strong Python proficiency (PyMC, Stan, NumPy, SciPy, PyTorch/JAX preferred)
• Experience implementing mathematical models that scale to high-dimensional data
• Ability to work with ML teams without being a “black-box ML” practitioner

Strongly Preferred (but Not Required)

• Experience in systems biology, genomics, transcriptomics, or proteomics
• Familiarity with biological pathway databases (KEGG, Reactome, SIGNOR, etc.)
• Prior work on regulatory-facing, interpretable models in life sciences
• Experience translating theory into production-grade inference pipelines