Data Scientist

The Barcelona Institute for Global Health (ISGlobal) is a cutting-edge institute addressing global public health challenges through research, translation into policy and education. ISGlobal has a broad portfolio in communicable and non-communicable diseases including environmental and climate determinants, and applies a multidisciplinary scientific approach ranging from the molecular to the population level. Research is organized in five programs: Climate, Air Pollution, Nature and Urban Health; Environment and Health over the Lifecourse; Global Viral and Bacterial Infections; Malaria and Neglected Parasitic Diseases and Maternal Child and Reproductive Health. ISGlobal is accredited with the Severo Ochoa distinction, a seal of excellence of the Spanish Science Ministry.

**WHAT WE ARE LOOKING FOR**:
**KEY RESPONSIBILITIES**:
Phase 1 - Retrospective Modeling and Data Integration
- Lead the curation, cleaning, and harmonization of large, multi-source biomedical datasets (clinical, histopathology, microbiology) derived from the CHAMPS, CaDMIA/CaDMIA-Plus, and São Paulo studies.
- Develop and implement AI-assisted data preprocessing pipelines and algorithms for automated data extraction, normalization, and validation.
- Conduct model validation and performance evaluation using gold-standard autopsy datasets (CDA), generating metrics such as sensitivity, specificity, kappa, and AUC.
- Collaborate with pathologists, microbiologists, and clinicians to interpret analytical results and integrate them into evidence-based diagnostic frameworks.
- Prepare technical documentation, visual reports, and scientific summaries to support protocol refinement and stakeholder communication.

Phase 2 - Prospective Validation and Multicenter Data Management
- Support the setup and harmonization of multicenter data collection systems during the prospective validation phase in Mozambique and Nepal.
- Implement automated data monitoring tools and consistency checks to ensure data integrity across sites.
- Participate in the analysis of within-case comparisons between simplified MITS (s-MITS), conventional MITS (c-MITS), and complete diagnostic autopsy (CDA) results.
- Contribute to the generation of data visualizations, dashboards, and analytical reports for internal and external dissemination.
- Collaborate in scientific publications, presentations, and consortium meetings (e.g., CODA, CHAMPS, and MITS Alliance).
- Provide technical mentorship and support to junior analysts, trainees, or partner site data teams as required.

**AUXILIAR TASKS**

This job description reflects the present requirements of the post but may evolve at any time in the future as duties and responsibilities change and/or develop providing there is appropriate consultation with the post-holder.

**SKILLS**:

- Solid understanding of biomedical data structures, including clinical, microbiological, and pathology datasets.
- Strong capacity for critical thinking and scientific reasoning, particularly in interpreting model outputs in biological and epidemiological contexts.
- Experience in data visualization and preparation of analytical reports for technical and non-technical audiences.
- Attention to detail and a rigorous approach to data quality assurance and documentation.
- Familiarity with data protection principles and good practices for managing sensitive health information.
- Ability to manage multiple tasks and deadlines in a dynamic, research-driven environment.
- Enthusiasm for working at the interface between data science, medicine, and global health, and commitment to ISGlobal’s mission to improve health equity worldwide.

The post holder will adhere to ISGlobal principles contained in **People management policy, including Equity, diversity and health safety**. The post holder will respect, and be accountable to ensure ISGlobal policies and procedures.

**TRAINING AND EXPERIENCE / QUALIFICATIONS**:
Required:

- PhD or MSc degree (with at least 3 years of relevant experience) in Bioinformatics, Computational Biology, Data Science, Biostatistics, or a related quantitative discipline.
- Proven experience in data analysis and modeling of biomedical or epidemiological datasets, including integration of heterogeneous data types (clinical, microbiological, histopathological, or genomic).
- Strong programming and analytical skills in Python and/or R, including proficiency with data wrangling, statistical modeling, and visualization packages.
- Demonstrated expertise in artificial intelligence and machine learning, including supervised and unsupervised modeling, feature selection, and model validation.
- Familiarity with version control and reproducible research tools (Git/GitHub, Jupyter, R Markdown, or similar).
- Strong organizational and problem-solving skills, with capacity to work independently and collaboratively in a multidisciplinary team.

Desirable:

- Previous experience in global health research, mortality surveillance, or clinical/pathology data