Scientist - Phenotypic Data
Genomics LondonEst. Est. GBP 50,000–70,000 / yearMid
Estimated range based on role, country and industry — not published by the company.
Key requirements
- Python
- Sql
- Machine Learning
- Nlp
Location: Oxford, London or RTP (Hybrid)
The Mission: Why We Exist
Genomics is a science-led transatlantic TechBio combining large-scale genetic and health data with proprietary analytics to accelerate drug discovery and advance predictive, preventative healthcare. We are united by a single vision to help people live longer healthier lives, using the power of genomics.
Genomics aims to help people live longer, healthier lives in two ways: super-charging drug discovery and development for novel treatments with our AI-enabled advanced genetic analytics platform, and by helping people understand their personal risk of common chronic diseases through polygenic risk scores - giving doctors and health systems the chance to get the right people into the right prevention, screening and treatment programmes at the right time.
Role Purpose
Genetic analyses are only as good as the definitions of disease they’re built on. Get the phenotype wrong and everything downstream wobbles.
As a Scientist in Clinical Informatics, you’ll define, curate and validate phenotypes from electronic health records and other clinical data. That gives our genetic discovery work and risk tools foundations that are accurate, clinically meaningful and reproducible. You’ll join the Life Sciences team, which tackles challenges across the therapeutic development cycle, from target discovery to patient stratification and finding new opportunities for existing drugs.
This is an early-career role in a highly collaborative team. You’ll be supported to develop your own ownership of analyses, and you’ll see your work reused across projects.
A Day in the Life
Developing and validating phenotype definitions for diseases with unmet need, using coded and clinical data (diagnoses, procedures, prescriptions, laboratory measurements) and standard terminologies, and assessing their accuracy and potential sources of bias
Harmonising and characterising phenotypic data across cohorts and health data sources, describing disease onset, progression and treatment patterns to identify relevant patient populations and biomarkers
Applying and evaluating risk stratification tools that combine polygenic risk scores with conventional clinical risk factors, to find the people most at risk and those most likely to benefit from particular therapies
Contributing reusable, version-controlled phenotyping code, pipelines and methods that feed into our technology
Working with genetic analysts, data engineers and clinical advisers, and presenting your findings to scientific and clinical audiences
Who You Are
You’ll bring:
A PhD in a quantitative discipline (epidemiology, biostatistics, health data science, medical informatics or genetics, for example)
Solid foundations in statistics (modelling, regression)
Experience with clinical or health data, such as electronic health records, cohort or registry data
Statistical programming in Python or R
Clear spoken and written communication, the ability to manage your own objectives and deliver on time, and a team-player mindset that puts your work in the bigger picture
Comfort with large, complex, imperfect clinical datasets, and a scientist’s instinct for questions of data quality and validity
It’s a bonus if you have:
Experience defining phenotypes from EHRs, using coded data or free text
Familiarity with clinical terminologies and data models (ICD-10, SNOMED CT, OPCS-4, OMOP, HPO)
An understanding of epidemiological study design and bias (confounding, selection bias, measurement error)
Knowledge of survival or longitudinal modelling and causal inference
Familiarity with statistical genetics (GWAS, polygenic risk scores), genetics and molecular biology
Experience applying NLP or machine learning to clinical text or structured health data
SQL and software engineering practices (Git and GitHub, testing, documentation, agentic coding, containerisation)
Reproducible pipelines (WDL, Snakemake or Nextflow) a
See your match score for this role.
Xecodai maps the interview stages and shows what is preventing a 95% match.